Computed tomography of the abdomen

Citation for published version (APA):

Martens, B. (2022). Computed tomography of the abdomen: from one size fits all to custom-made.
[Doctoral Thesis, Maastricht University]. Ridderprint. https://doi.org/10.26481/dis.20220317bm

Document status and date:

Published: 01/01/2022

DOI:
10.26481/dis.20220317bm

Document Version:
Publisher's PDF, also known as Version of record

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can
be important differences between the submitted version and the official published version of record.
People interested in the research are advised to contact the author for the final version of the publication,
or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page
numbers.
Link to publication

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these
rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above,
please follow below link for the End User Agreement:
www.umlib.nl/taverne-license

Take down policy

If you believe that this document breaches copyright please contact us at:
[email protected]
providing details and we will investigate your claim.

Download date: .
 01/02/2022 21:06:00
Computed tomography of the abdomen: From one size fits all to custom-made Bibi Martens

Omslag_BM_1.indd 2-3
Vol_BM_productie.indd 1 02/02/2022 15:22:44
 © Copyright by Bibi Martens, Maastricht 2022
ISBN: 978-94-6458-063-1.

No parts of this publication may be reproduced, stored in a retrieval system, or

transmitted in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without permission in writing from the author.

Cover photo: Becky Pruitt

Models: Horse eye Jacks, Bonaire
Cover design: Mickel Zelders & Bibi Martens
Printing: Ridderprint, www.ridderprint.nl
Layout and design: Robin Weijland, persoonlijkproefschrift.nl

Vol_BM_productie.indd 2 02/02/2022 15:22:44

Vol_BM_productie.indd 3 02/02/2022 15:22:44
Vol_BM_productie.indd 4 02/02/2022 15:22:44
 CONTENTS

CHAPTER 1 General introduction 7

CHAPTER 2 Individually body weight–adapted contrast media 23

application in computed tomography imaging of the
liver at 90 kVp

CHAPTER 3 A solution for homogeneous liver enhancement in 41

computed tomography: results from the COMpLEx Trial

CHAPTER 4 Finding the optimal tube current and iterative 61

reconstruction strength in liver imaging; two needles
in one haystack

CHAPTER 5 Individualized scan protocols in abdominal computed 79

tomography: radiation versus contrast media dose
optimization

CHAPTER 6 Influence of contrast material temperature on patient 97

comfort and image quality in computed tomography of
the abdomen (CATCHY): a randomized controlled trial

CHAPTER 7 Tailoring contrast media protocols to varying tube 115

voltages in vascular and parenchymal CT imaging. The
10-to-10 rule

CHAPTER 8 Book chapter in “Artificial Intelligence in Cardiothoracic 129

Imaging” title: Artificial intelligence based contrast
medium optimization

CHAPTER 9 General discussion 145

CHAPTER 10 Summary 159

Nederlandse samenvatting

CHAPTER 11 Scientific impact 167

CHAPTER 12 Dankwoord 173

List of publications
Curriculum vitae

Vol_BM_productie.indd 5 02/02/2022 15:22:44

Vol_BM_productie.indd 6 02/02/2022 15:22:46
 CHAPTER
1
General introduction

Vol_BM_productie.indd 7 02/02/2022 15:22:47

 Chapter 1

Millions of Computed Tomography (CT) scans are performed worldwide each

year, with a large variety of scan and contrast media (CM) protocols (1). Non-
invasive CT is the first-choice imaging modality for various clinical questions,
such as oncological (follow-up) and infectious or vascular evaluation. Over the
years faster scanners with high spatial and temporal resolutions have made
CT a workhorse of daily clinical practice (2).

CT is associated with ionizing radiation, which in case of high exposure may

lead to an increase in the lifetime attributable cancer risk (3, 4). This is the
main reason for universal application of the “as low as reasonably achievable”
(ALARA) principle (5). A large number of studies have focused on decreasing
radiation dose (6-13). Radiation dose can be reduced by decreasing tube current
and/or tube voltage. Initially, only tube current reduction was available, but
presently tube current can be freely modulated. In addition, the development
of more powerful tubes has enabled lower tube voltage, albeit limited to only
few settings.

In the past, ‘one size fits all’ protocols were used for both radiation and CM
injection. In other words, all patients were scanned with the same tube
current and tube voltage, regardless of clinical question or body composition.
At present, both parameters are determined before scanning, based on the
scout view, and further adjusted during the examination to optimize radiation
dose for each individual patient and body part. Automated tube current
modulation (ATCM) and automated tube voltage selection (ATVS) techniques,
currently present on the majority of newer scanners, have made it easier to
individualize scan protocols (2, 14). These do not require any intervention from
the radiographer and are thus less time consuming.

In addition to ATCM and ATVS, iterative reconstruction (IR) techniques are

commonly used, primarily to reduce radiation dose. IR reduces image noise
by repetitive calculation during reconstruction. Repetition is stopped after
completion of a predefined number of cycles, or when the difference between
two IR steps has become smaller than a predefined value (15).

A user set image quality for a standard patient is the reference basis for the
image quality level in each patient. When using ATCM and ATVS, user set image

Vol_BM_productie.indd 8 02/02/2022 15:22:47

 General introduction

quality is determined by reference tube current (mAsref ) and reference tube

voltage (kVref ). Furthermore, increasing IR strength is related to a decrease in
image noise. Various mAsref and kVref values and IR strengths are used in daily
clinical practice (16-19). The main goal is to define optimal settings for reaching
diagnostic image quality at lowest possible radiation dose.

In contrast-enhanced CT scans CM injection protocols can be adapted to

achieve diagnostic image quality at lowest possible CM dosage. The most 1
decisive parameter to base CM injection protocols on is iodine delivery rate
(IDR) for vascular studies, and total CM volume for parenchymal studies (20).
IDR (gI/s) can be calculated using CM concentration (in mg/ml) multiplied by the
CM injection flow rate (in ml/s). In most hospitals a single CM concentration is
used, and it is therefore straightforward to adapt flow rate to modify IDR (20).

Figure 1. When contrast (CM) volume and flow rate are kept constant, attenuation of
the liver parenchyma is determined by body weight.

For abdominal imaging, a ‘one size fits all’ CM injection protocol is used in many
centres, i.e. all patients receive the same amount of CM. As a consequence,
CM dose is only optimal in a small selection of the patient population (figure
1): patients with lower body weight may receive excess total CM volume which

Vol_BM_productie.indd 9 02/02/2022 15:22:47

 Chapter 1

may lead to very high attenuation and artifacts, whereas heavier patients
may receive an insufficient amount of CM for adequate attenuation of the
organs. In recent years, different body size indices have been proposed to
individualize CM injection protocols. Of these, total body weight is the easiest,
measured for example in the scanner room using a calibrated weighing scale.
Lean body weight (LBW, i.e. total body weight minus fat), and body surface are
(BSA), require more complicated calculations (21-27). Previous studies have
shown that body weight adapted CM protocols result in more homogeneous
enhancement of both pulmonary and coronary arteries, compared to a ‘one
size fits all’ protocol (28, 29), which may be similarly applicable to abdominal
imaging. Furthermore, individualized protocols for coronary and pulmonary
arteries resulted in an overall reduction in CM volume for patients with lower
body weight (28, 29). A relatively easy way to individualize CM injection protocols
is using dedicated CM injection software based on the non-linear relationship
between body weight and scan duration (28, 29). The flow rate then depends
on the total amount of CM required based on patient body weight (28, 30).

Radiation dose parameters and CM injection protocols cannot be seen as

separate entities. This is because tube voltage influences attenuation of iodine
(31). When the tube voltage reaches the 33 keV k-edge of iodine, the chance
an electron will be dislodged from the k-shell increases, with an increase in
photo-electric energies as a result. Therefore, attenuation of iodine increases
with decreasing tube voltage and consequently, influencing the one may
require adjusting the other. Figure 2 is an example of a patient with a small
enhancing kidney lesion, who undergoes regular CT scans for follow-up. The
figure shows attenuation of the lesion on an unenhanced CT scan and after
injection of iodinated CM. The large differences in attenuation between scans
can be attributed to both the lower tube voltage and the larger CM volume
used in the second scan. The diagnosis of a malignant kidney lesion depends
on CM enhancement (> 15 Hounsfield Units [HU] enhancement in arterial phase
compared to the unenhanced phase is suggestive of a renal cell carcinoma),
which illustrates the importance of taking both radiation and CM injection
protocol into account when acquiring and reading such images (32, 33).

10

Vol_BM_productie.indd 10 02/02/2022 15:22:47

 General introduction

Many studies focus on optimizing either radiation dose or CM injection

protocols, which, due to the strong connection between the two, is of limited
use without a rule of thumb to match them (6-10, 28, 29, 31).

Figure 2. Images of two repeat scans. (A,B,C and D,E,F) of an enhancing lesion in the
right kidney. The two scans were done a year apart and using different scan parameters.
Because 100 kV is closer to the 33 keV k-edge of iodine, attenuation of iodine is in-
creased in the second scan. It is surprising therefore that a larger CM volume was used.
1

An incentive to such a rule was provided for vascular studies by Kok et al. (34).
They used a circulation phantom to demonstrate that when a tube voltage
reduction of 120 to 100 kV was accompanied by a 12 % reduction in CM dose,
sufficient attenuation of the coronary arteries was achieved (> 325 HU). These
results were confirmed in a small patient group (34). The hypothesis may be
extended to parenchymal studies. Based on the literature, a tube voltage
reduction of 10 kV accompanied by a 10 % reduction in CM volume may result
in homogeneous enhancement of the liver, regardless of tube voltage used
(23, 35). Even better would be the use of a dosing factor in g I/kg body weight

11

Vol_BM_productie.indd 11 02/02/2022 15:22:47

 Chapter 1

instead of CM volume: a 10 kV reduction accompanied by a 10 % decrease

in dosing factor and vice versa (35). Such a rule of thumb has not yet been
evaluated in a clinical setting, however.

There are pitfalls to reducing radiation and CM dose. Excessive radiation dose
reduction may lead to increased image noise and result in a non-diagnostic scan
(36). At best a non-diagnostic scan may need to be repeated, at worst it may
result in underdiagnosis. Small liver lesions not visible due to image noise, for
example, could have a large impact on patient treatment and survival. Similarly,
a large number of clinical questions require the use of intravenous iodinated
CM. Controversy remains on whether the post-contrast acute reduction in
renal function sometimes seen reflects kidney injury caused by iodinated CM
(37-40). Regardless, lowering CM dose is preferable especially in patients with
reduced renal function, and there is simply no reason to give any patient more
CM than necessary for diagnostic purposes. The question is, how low can we
go without compromising diagnostic image quality?

Diagnostic image quality seems straightforward, while in truth it is quite a

difficult concept. Image quality can be assessed objectively or subjectively.
Objective image quality encompasses attenuation in HU, the signal to noise
ratio (SNR) and the contrast to noise ratio (CNR). SNR is calculated by dividing
the attenuation of the target organ by its standard deviation (SD), while CNR is
the difference between the attenuation of the organ and the paraspinal muscle,
divided by the SD of the same muscle (8, 29, 41, 42). These measurements are
quite easy but thresholds are not always well defined: for abdominal imaging
a wide variety of acceptable values are found in literature. For subjective
image quality the picture is even less clear. The introduction of IR techniques
in daily clinical practice has tremendously changed subjective image quality
assessment (43), and no clear-cut values exist (9, 10, 44-47). Subjective image
quality is often rated on a 5-point Likert scale (e.g. 1, excellent; 2, good; 3,
moderate; 4,poor; 5, very poor) (48), but it remains a subjective score influenced
by the individual radiologist’s preference. Dividing the subjective image quality
parameter into smaller units (e.g. noise, contrast, lesion detectability), may
improve uniformity. However, in practice image quality is more than just
the sum of those parameters. It is clear that both objective and subjective

12

Vol_BM_productie.indd 12 02/02/2022 15:22:47

 General introduction

parameters have their limitations, but no better alternative has been proposed
yet, and at this time we have to work with what we’ve got.

Image quality being so difficult to capture with one reliable parameter, studies
would profit from intra-patient comparisons. However, for obvious ethical and
ALARA-related principles, this is not possible. In the current era, where artificial
intelligence (AI) techniques are emerging, reconstruction software might be of
assistance. Using dedicated post-processing software able to mimic lower tube 1
current by inserting noise to the CT image, studies have shown that a radiation
dose reduction of 41 to 84 % was possible in CT angiography of various vascular
structures in head and neck, without compromising diagnostic image quality
(49, 50). No studies have done pairwise comparisons of different radiation
doses and IR strengths in abdominal imaging within the same patient, and
whether a dose reduction still leads to sufficient image quality in this setting
begs to be investigated.

Age and kidney function may affect radiation dose and contrast volume protocol
considerations. In a younger patient, reducing radiation dose is important to
decrease lifetime attributable cancer risk (3-5), whereas in the older population,
where reduced kidney function is more common, a CM dose reduction may
be more important (3-5, 37-40). In ATVS, settings (slider level) can be adjusted
to optimize either radiation dose or CM (51). In daily clinical practice, the slider
is set according to the type of CT performed (e.g. vascular, parenchymal or
unenhanced). For each slider setting, a user set reference image quality is
specified, based on the CNR. In vascular studies, CM is leading and in general
more noise is accepted so that radiation dose can be lower, and a slider position
11 is chosen. For parenchymal studies a balance between attenuation and
noise is preferred and the slider is set at position 7. Unenhanced scans use
position 3, in which the CNR is based solely on the fat-water contrast (51).
Adapting either radiation dose (by changing tube voltage) or CM volume based
on age and/or kidney function may be preferable in further optimizing scan and
CM injection protocols. Euler et al. showed that this was feasible in vascular CT’s,
resulting in comparable objective and subjective image quality (51). However,
vascular studies have other requirements than parenchymal studies, and slider
position manipulation needs to be explored in the setting of abdominal imaging.

13

Vol_BM_productie.indd 13 02/02/2022 15:22:48

 Chapter 1

Whereas scan and CM injection protocols have been extensively studied, one
basic parameter is still under debate: CM temperature. The European Society of
Urogenital Radiology (ESUR) and American College of Radiology (ACR) guidelines
differ in their recommendations regarding the necessity to prewarm CM prior
to injection: ESUR advises standard pre-warming whereas ACR states warming
may only be helpful in certain specific circumstances. Increasing CM iodine
concentration increases its viscosity, pre-warming CM decreases its viscosity
(52-56). It was long thought that injecting CM at a high flow rate (> 6 ml/s)
resulted in decreased patient comfort and increased risk of CM extravasations.
Patient comfort is important not only for the patient but also for the procedure.
An uncomfortable patient might start moving or shivering, breathe more
quickly and/or have a faster heartbeat, all of which negatively influence image
quality. The EICAR trial showed that injecting iodinated CM with a flow rate as
high as 8.3 ml/s is safe and does not increase the risk of CM extravasations
and/or pain when CM is pre-warmed to 37° C (99° F) (52). On the other hand,
Davenport et al. suggest that pre-warming CM is not necessary for low iodine
concentrations. In their study a total of 12.682 pre-warmed injections and
12.138 injections at room temperature were retrospectively evaluated with
regard to CM extravasations and adverse events (57). The results were not
conclusive: although it is safe to inject CM at room temperature, it could result
in a decreased patient comfort and pain. With regard to the individualization,
optimalization and efficiency of daily clinical practice, it would be valuable to
know whether pre-warming CM is effective in reducing CM adverse events and
increasing patient comfort.

Automated systems play a large role in radiation dose optimization (ATCM and
ATVS), but CM administration is still often a manual action. To easily, quickly
and reliably individualize both radiation and CM, it would be preferable that
CM injection protocols be built into the system, linked to the scanner. In that
respect, AI might be helpful. AI could help selecting the optimal CM injection
protocol for an individual patient on a particular scanner and for a specific scan
indication, improving patient care and workflow. Even factors such as patient
anxiety and difficulty gaining venous access, which may extend the duration
of a scan, could be taken into account for patients having a CT appointment
on a regular basis (e.g. oncological follow-up), improving the lead time (58).

14

Vol_BM_productie.indd 14 02/02/2022 15:22:48

 General introduction

Ultimately AI might be able to help in reducing CM volume, scan time, and

radiation dose (59).

The aim of this thesis is to optimize and individualize CT protocols for abdominal
imaging. The ultimate goal is to provide a combined protocol tailored to both
individual patient parameters and clinical question.

OUTLINE OF THIS THESIS 1

In Chapter 2 image quality (objective and subjective) is compared between

body weight-adapted and standard CM-volume injection protocols.

In Chapter 3 the relationship between tube voltages and CM injection protocols

is studied for abdominal portal venous phase imaging in a randomized
controlled trial. A reduction in tube voltage leads to an increase in attenuation
of iodine, and therefore may enable a reduction in CM dose.

In Chapter 4, optimal reference tube current and IR strength for abdominal

imaging is investigated using dedicated simulation software for offline
reconstruction of CT images based on raw-data sets.

Where previous chapters evaluate individualized scan protocols based on body

weight and tube voltage, Chapter 5 details a feasibility study in which radiation
and CM injection protocols are adjusted to take age and renal function into
account.

Chapter 6 reports the results of a randomized controlled trial comparing

patient comfort and image quality between protocols using pre-warmed or
room temperature CM.

Chapter 7 provides a summary of the most important parameters in both

parenchymal and vascular CT imaging. An easy-to-use rule of thumb is
proposed that can be applied to individualize CM injection protocols based on
both patient body weight and tube voltage.

15

Vol_BM_productie.indd 15 02/02/2022 15:22:48

 Chapter 1

In Chapter 8 a comprehensive overview of recent developments in optimizing

scan and CM injection protocols is given. With the arrival of AI, new possibilities
arise for further radiation and CM dose reductions in the near future. Although
the chapter focuses on the coronary arteries, proposed techniques may be
applicable to a much broader field.

Chapter 9 contains a general discussion of all previous chapters and future

perspectives.

References

1. Smith-Bindman R, Wang Y, Chu P, Chung R, Einstein AJ, Balcombe J, et al. International

variation in radiation dose for computed tomography examinations: prospective
cohort study. BMJ. 2019;364:k4931.
2. Lell MM, Wildberger JE, Alkadhi H, Damilakis J, Kachelriess M. Evolution in Computed
Tomography: The Battle for Speed and Dose. Invest Radiol. 2015;50(9):629-44.
3. The 2007 Recommendations of the International Commission on Radiological
Protection. ICRP publication 103. Annals of the ICRP. 2007;37(2-4):1-332.
4. Board of Radiation Effects Research Division on Earth and Life Sciences National
Research Council of the National Academies. Health Risks from Exposure to Low
Levels of Ionizing Radiation: BEIR VII, Phase 2. Washington (DC)2006.
5. Barrett B, Stiles M, Patterson J. Radiation risks: critical analysis and commentary.
Prev Med. 2012;54(3-4):280-2.
6. Holmquist F, Soderberg M, Nyman U, Falt T, Siemund R, Geijer M. 80-kVp hepatic
CT to reduce contrast medium dose in azotemic patients: a feasibility study. Acta
Radiol. 2020;61(4):441-9.
7. Leyendecker P, Faucher V, Labani A, Noblet V, Lefebvre F, Magotteaux P, et al.
Prospective evaluation of ultra-low-dose contrast-enhanced 100-kV abdominal
computed tomography with tin filter: effect on radiation dose reduction and image
quality with a third-generation dual-source CT system. Eur Radiol. 2019;29(4):2107-
16.
8. Shuman WP, Chan KT, Busey JM, Mitsumori LM, Choi E, Koprowicz KM, et al.
Standard and reduced radiation dose liver CT images: adaptive statistical iterative
reconstruction versus model-based iterative reconstruction-comparison of findings
and image quality. Radiology. 2014;273(3):793-800.
9. Goshima S, Kanematsu M, Noda Y, Kawai N, Kawada H, Ono H, et al. Minimally
Required Iodine Dose for the Detection of Hypervascular Hepatocellular Carcinoma
on 80-kVp CT. AJR Am J Roentgenol. 2016;206(3):518-25.

16

Vol_BM_productie.indd 16 02/02/2022 15:22:48

 General introduction

10. Nakamoto A, Kim T, Hori M, Onishi H, Tsuboyama T, Sakane M, et al. Clinical

evaluation of image quality and radiation dose reduction in upper abdominal
computed tomography using model-based iterative reconstruction; comparison
with filtered back projection and adaptive statistical iterative reconstruction. Eur J
Radiol. 2015;84(9):1715-23.
11. Brehmer K, Brismar TB, Morsbach F, Svensson A, Stal P, Tzortzakakis A, et al. Triple
Arterial Phase CT of the Liver with Radiation Dose Equivalent to That of Single
Arterial Phase CT: Initial Experience. Radiology. 2018;289(1):111-8.
12. Buls N, Van Gompel G, Van Cauteren T, Nieboer K, Willekens I, Verfaillie G, et al. 1
Contrast agent and radiation dose reduction in abdominal CT by a combination
of low tube voltage and advanced image reconstruction algorithms. Eur Radiol.
2015;25(4):1023-31.
13. Nakaura T, Nakamura S, Maruyama N, Funama Y, Awai K, Harada K, et al. Low
contrast agent and radiation dose protocol for hepatic dynamic CT of thin adults at
256-detector row CT: effect of low tube voltage and hybrid iterative reconstruction
algorithm on image quality. Radiology. 2012;264(2):445-54.
14. Kaza RK, Platt JF, Goodsitt MM, Al-Hawary MM, Maturen KE, Wasnik AP, et al.
Emerging techniques for dose optimization in abdominal CT. Radiographics : a
review publication of the Radiological Society of North America, Inc. 2014;34(1):4-17.
15. Stiller W. Basics of iterative reconstruction methods in computed tomography: a
vendor-independent overview. Eur J Radiol. 2018;109:147-54.
16. Hardie AD, Nelson RM, Egbert R, Rieter WJ, Tipnis SV. What is the preferred strength
setting of the sinogram-affirmed iterative reconstruction algorithm in abdominal
CT imaging? Radiol Phys Technol. 2015;8(1):60-3.
17. Choy S, Parhar D, Lian K, Schmiedeskamp H, Louis L, O’Connell T, et al. Comparison of
image noise and image quality between full-dose abdominal computed tomography
scans reconstructed with weighted filtered back projection and half-dose scans
reconstructed with improved sinogram-affirmed iterative reconstruction (SAFIRE*).
Abdominal Radiology. 2018;44.
18. Wang R, Schoepf UJ, Wu R, Nance JW, Jr., Lv B, Yang H, et al. Diagnostic accuracy
of coronary CT angiography: comparison of filtered back projection and iterative
reconstruction with different strengths. J Comput Assist Tomogr. 2014;38(2):179-84.
19. Kataria B, Nilsson Althen J, Smedby O, Persson A, Sokjer H, Sandborg M. Assessment
of image quality in abdominal computed tomography: effect of model-based iterative
reconstruction, multi-planar reconstruction and slice thickness on potential dose
reduction. Eur J Radiol. 2020;122:108703.
20. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
21. Awai K, Kanematsu M, Kim T, Ichikawa T, Nakamura Y, Nakamoto A, et al. The Optimal
Body Size Index with Which to Determine Iodine Dose for Hepatic Dynamic CT: A
Prospective Multicenter Study. Radiology. 2016;278(3):773-81.

17

Vol_BM_productie.indd 17 02/02/2022 15:22:48

 Chapter 1

22. Kondo H, Kanematsu M, Goshima S, Watanabe H, Kawada H, Moriyama N, et al.

Body size indices to determine iodine mass with contrast-enhanced multi-detector
computed tomography of the upper abdomen: does body surface area outperform
total body weight or lean body weight? Eur Radiol. 2013;23(7):1855-61.
23. Heiken JP, Brink JA, McClennan BL, Sagel SS, Crowe TM, Gaines MV. Dynamic
incremental CT: effect of volume and concentration of contrast material and patient
weight on hepatic enhancement. Radiology. 1995;195(2):353-7.
24. Bae KT, Shah AJ, Shang SS, Wang JH, Chang S, Kanematsu M, et al. Aortic and hepatic
contrast enhancement with abdominal 64-MDCT in pediatric patients: effect of body
weight and iodine dose. AJR Am J Roentgenol. 2008;191(5):1589-94.
25. Kondo H, Kanematsu M, Goshima S, Watanabe H, Onozuka M, Moriyama N, et al.
Aortic and hepatic enhancement at multidetector CT: evaluation of optimal iodine
dose determined by lean body weight. Eur J Radiol. 2011;80(3):e273-7.
26. Kondo H, Kanematsu M, Goshima S, Tomita Y, Kim MJ, Moriyama N, et al. Body
size indexes for optimizing iodine dose for aortic and hepatic enhancement at
multidetector CT: comparison of total body weight, lean body weight, and blood
volume. Radiology. 2010;254(1):163-9.
27. Matsumoto Y, Masuda T, Sato T, Arataki K, Nakamura Y, Tatsugami F, et al. Contrast
Material Injection Protocol With the Dose Determined According to Lean Body
Weight at Hepatic Dynamic Computed Tomography: Comparison Among Patients
With Different Body Mass Indices. Journal of computer assisted tomography.
2019;43(5):736-40.
28. Mihl C, Kok M, Altintas S, Kietselaer BL, Turek J, Wildberger JE, et al. Evaluation
of individually body weight adapted contrast media injection in coronary CT-
angiography. Eur J Radiol. 2016;85(4):830-6.
29. Hendriks BM, Kok M, Mihl C, Bekkers SC, Wildberger JE, Das M. Individually
tailored contrast enhancement in CT pulmonary angiography. Br J Radiol.
2016;89(1061):20150850.
30. Seifarth H, Puesken M, Kalafut JF, Wienbeck S, Wessling J, Maintz D, et al. Introduction
of an individually optimized protocol for the injection of contrast medium for
coronary CT angiography. Eur Radiol. 2009;19(10):2373-82.
31. Fleischmann U, Pietsch H, Korporaal JG, Flohr TG, Uder M, Jost G, et al. Impact of
Contrast Media Concentration on Low-Kilovolt Computed Tomography Angiography:
A Systematic Preclinical Approach. Invest Radiol. 2018;53(5):264-70.
32. Dyer R, DiSantis DJ, McClennan BL. Simplified imaging approach for evaluation of
the solid renal mass in adults. Radiology. 2008;247(2):331-43.
33. Kang SK, Huang WC, Pandharipande PV, Chandarana H. Solid renal masses: what
the numbers tell us. AJR Am J Roentgenol. 2014;202(6):1196-206.

18

Vol_BM_productie.indd 18 02/02/2022 15:22:48

 General introduction

34. Kok M, Mihl C, Hendriks BM, Altintas S, Kietselaer BL, Wildberger JE, et al. Optimizing
contrast media application in coronary CT angiography at lower tube voltage:
evaluation in a circulation phantom and sixty patients. Eur J Radiol. 2016;85(6):1068-
74.
35. Canstein C, Korporaal JG. Reduction of contrast agent dose at low kV settings. In:
Siemens Wp, editor. Forchheim, Germany2015.
36. Geyer LL, Schoepf UJ, Meinel FG, Nance JW, Jr., Bastarrika G, Leipsic JA, et al. State
of the Art: Iterative CT Reconstruction Techniques. Radiology. 2015;276(2):339-57.
37. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT. Hospital-acquired renal 1
insufficiency: a prospective study. Am J Med. 1983;74(2):243-8.
38. Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin
Pract. 2012;120(4):c179-84.
39. McDonald RJ, McDonald JS, Bida JP, Carter RE, Fleming CJ, Misra S, et al. Intravenous
contrast material-induced nephropathy: causal or coincident phenomenon?
Radiology. 2013;267(1):106-18.
40. Nijssen EC, Rennenberg RJ, Nelemans PJ, Essers BA, Janssen MM, Vermeeren
MA, et al. Prophylactic hydration to protect renal function from intravascular
iodinated contrast material in patients at high risk of contrast-induced nephropathy
(AMACING): a prospective, randomised, phase 3, controlled, open-label, non-
inferiority trial. Lancet. 2017;389(10076):1312-22.
41. Szucs-Farkas Z, Strautz T, Patak MA, Kurmann L, Vock P, Schindera ST. Is body weight
the most appropriate criterion to select patients eligible for low-dose pulmonary
CT angiography? Analysis of objective and subjective image quality at 80 kVp in 100
patients. Eur Radiol. 2009;19(8):1914-22.
42. Song JS, Lee JM, Sohn JY, Yoon JH, Han JK, Choi BI. Hybrid iterative reconstruction
technique for liver CT scans for image noise reduction and image quality
improvement: evaluation of the optimal iterative reconstruction strengths. Radiol
Med. 2015;120(3):259-67.
43. Willemink MJ, Leiner T, de Jong PA, de Heer LM, Nievelstein RA, Schilham AM, et al.
Iterative reconstruction techniques for computed tomography part 2: initial results
in dose reduction and image quality. Eur Radiol. 2013;23(6):1632-42.
44. Chen CY, Hsu JS, Jaw TS, Kuo YT, Wu DC, Lee CH, et al. Lowering radiation dose during
dedicated colorectal cancer MDCT: comparison of low tube voltage and sinogram-
affirmed iterative reconstruction at 80 kVp versus blended dual-energy images in a
population of patients with low body mass index. Abdom Imaging. 2015;40(7):2867-
76.
45. Goshima S, Kanematsu M, Noda Y, Kondo H, Watanabe H, Kawada H, et al.
Determination of optimal intravenous contrast agent iodine dose for the detection
of liver metastasis at 80-kVp CT. Eur Radiol. 2014;24(8):1853-9.

19

Vol_BM_productie.indd 19 02/02/2022 15:22:48

 Chapter 1

46. Scholtz JE, Wichmann JL, Husers K, Beeres M, Nour-Eldin NE, Frellesen C, et al.
Automated tube voltage adaptation in combination with advanced modeled
iterative reconstruction in thoracoabdominal third-generation 192-slice dual-source
computed tomography: effects on image quality and radiation dose. Acad Radiol.
2015;22(9):1081-7.
47. Kanematsu M, Kondo H, Miyoshi T, Goshima S, Noda Y, Tanahashi Y, et al. Whole-body
CT with high heat-capacity X-ray tube and automated tube current modulation--
effect of tube current limitation on contrast enhancement, image quality and
radiation dose. Eur J Radiol. 2015;84(5):877-83.
48. Caruso D, De Santis D, Rivosecchi F, Zerunian M, Panvini N, Montesano M, et al.
Lean body weight-tailored iodinated contrast injection in obese patient: boer versus
james formula. Biomed Res Int. 2018;2018:8521893.
49. Ellmann S, Kammerer F, Brand M, Allmendinger T, May MS, Uder M, et al. A Novel
Pairwise Comparison-Based Method to Determine Radiation Dose Reduction
Potentials of Iterative Reconstruction Algorithms, Exemplified Through Circle of
Willis Computed Tomography Angiography. Invest Radiol. 2016;51(5):331-9.
50. Kramer M, Ellmann S, Allmendinger T, Eller A, Kammerer F, May MS, et al.
Computed Tomography Angiography of Carotid Arteries and Vertebrobasilar
System: A Simulation Study for Radiation Dose Reduction. Medicine (Baltimore).
2015;94(26):e1058.
51. Euler A, Taslimi T, Eberhard M, Kobe A, Reeve K, Zimmermann A, et al. Computed
Tomography Angiography of the Aorta-Optimization of Automatic Tube Voltage
Selection Settings to Reduce Radiation Dose or Contrast Medium in a Prospective
Randomized Trial. Invest Radiol. 2021;56(5):283-91.
52. Kok M, Mihl C, Hendriks BM, Altintas S, Eijsvoogel NG, Kietselaer BL, et al. Patient
comfort during contrast media injection in coronary computed tomographic
angiography using varying contrast media concentrations and flow rates: results
from the EICAR trial. Invest Radiol. 2016;51(12):810-5.
53. Kok M, Mihl C, Mingels AA, Kietselaer BL, Muhlenbruch G, Seehofnerova A, et al.
Influence of contrast media viscosity and temperature on injection pressure in
computed tomographic angiography: a phantom study. Invest Radiol. 2014;49(4):217-
23.
54. Roth R, Akin M, Deligonul U, Kern MJ. Influence of radiographic contrast media
viscosity to flow through coronary angiographic catheters. Cathet Cardiovasc Diagn.
1991;22(4):290-4.
55. American College of Radiology. Manual On Contrast Media: 2021 2021 [Available
from: https://www.acr.org/-/media/ACR/Files/Clinical-Resources/Contrast_Media.
pdf.
56. European Society of Urogenital Radiology. ESUR guidelines on contrast agents
European Society of Urogenital Radiology 10.0 2018 [Available from: http://www.
esur.org/fileadmin/content/2019/ESUR_Guidelines_10.0_Final_Version.pdf.

20

Vol_BM_productie.indd 20 02/02/2022 15:22:48

 General introduction

57. Davenport MS, Wang CL, Bashir MR, Neville AM, Paulson EK. Rate of contrast
material extravasations and allergic-like reactions: effect of extrinsic warming
of low-osmolality iodinated CT contrast material to 37 degrees C. Radiology.
2012;262(2):475-84.
58. Spyropoulos CD. AI planning and scheduling in the medical hospital environment.
Artif Intell Med. 2000;20(2):101-11.
59. Eberhard M, Alkadhi H. Machine Learning and Deep Neural Networks: Applications in
Patient and Scan Preparation, Contrast Medium, and Radiation Dose Optimization.
J Thorac Imaging. 2020;35 Suppl 1:S17-S20. 1

21

Vol_BM_productie.indd 21 02/02/2022 15:22:48

Vol_BM_productie.indd 22 02/02/2022 15:22:50
 CHAPTER
2
Individually body weight–adapted
contrast media application in
computed tomography imaging of
the liver at 90 kVp

B. Martens, B.M.F. Hendriks, N.G. Eijsvoogel, J.E. Wildberger, C. Mihl

Published in:
Investigative Radiology 2019; 54(3):177-182

Vol_BM_productie.indd 23 02/02/2022 15:22:51

 Chapter 2

Abstract

Objectives: The aim of the present study was to evaluate the attenuation and
image quality (IQ) of a body weight adapted contrast media (CM) protocol
compared to a fixed injection protocol in computed tomography (CT) of the
liver at 90 kV.

Materials and Methods: 199 consecutive patients referred for abdominal CT

imaging in portal venous phase were included. Group 1 (N =100) received a
fixed CM dose with a total iodine load (TIL) of 33 g I at a flow rate of 3.5 ml/s,
resulting in an iodine delivery rate (IDR) of 1.05 g I/s. Group 2 (N = 99) received
a body weight adapted CM protocol with a dosing factor of 0.4 g I/kg with a
subsequent TIL adapted to the patients’ weight. Injection time of 30 s was kept
identical for all patients. Therefore, flow rate and IDR changed with different
body weight. Patients were divided into three weight categories; ≤ 70 kg, 71
- 85 kg and ≥ 86 kg. Attenuation (HU) in three segments of the liver, signal-
to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were used to evaluate
objective IQ. Subjective IQ was assessed by a 5-point Likert scale. Differences
between groups were statistically analysed (p < 0.05 was considered statistically
significant).

Results: No significant differences in baseline characteristics were found

between groups. The CM volume and TIL differed significantly between groups
(p < 0.01), with mean values in group 1 of 110 ml and 33 g I; and in group 2
of 104.1 ± 21.2 ml and 31.2 ± 6.3 g I, respectively. Flow rate and IDR were
not significantly different between groups (p > 0.05). Body weight adapted
protocolling led to more homogeneous enhancement of the liver parenchyma
compared to a fixed protocol with a mean enhancement per weight category in
group 2 of; 126.5 ± 15.8; 128.2 ± 15.3 and 122.7 ± 21.2 HU, compared to 139.9
± 21.4; 124.6 ± 24.8 and 116.2 ± 17.8 HU in group 1, respectively.

Conclusions: Body weight adapted CM injection protocols result in more

homogeneous enhancement of the liver parenchyma at 90 kV in comparison
to a fixed CM volume with comparable objective and subjective IQ, while overall
CM volume can be safely reduced in more than half of patients.

24

Vol_BM_productie.indd 24 02/02/2022 15:22:51

 Body Weight–Adapted CM Application in CT Imaging

Keywords

Multidetector Computed Tomography; Diagnostic Imaging; Liver; Radiation

Dosage; Contrast Media

Introduction

Contrast enhanced computed tomography (CT) is frequently used to detect

liver lesions and to (sub-)classify liver tumours (1-4). Studies show that hepatic
enhancement of ≥ 50 HU is considered necessary to ensure an appropriate 2
visibility of low-attenuating lesions (5-9). Usually, a standard fixed contrast
media (CM) volume is used independent of specific patient characteristics such
as height, weight, liver status (e.g. cirrhosis and steatosis) and cardiac function
(6, 10, 11). All these factors have a direct influence on liver enhancement and,
as a result, on lesion visibility. A fixed CM dose therefore results in reduced
attenuation levels in the liver in heavier weighted patients, in comparison to
patients with a lower total body weight (TBW) (12). As a consequence, this might
even lead to scans with an insufficient attenuation resulting in a non-diagnostic
CT-scan (12). Alternatively, patients with a low TBW might receive more CM than
necessary for sufficient liver attenuation, which is not preferable either (13).

Dedicated CM injection software (P3T TM [Bayer Healthcare, Berlin, Germany])

individualizes CM application for each patient based on body weight and the
linear relationship between weight and flow rate (ml/s). As a result, the flow
rate and the resulting iodine delivery rate (IDR) differ with a changing TBW.
Injection time is constant for all patients. A higher TBW therefore, results in
a higher flow rate, with a subsequent increase in total iodine load (TIL). The
advantage of an individualized CM injection protocol over a fixed CM injection
protocol has already been established in various CT angiography (CTA) studies
(14-18). The effects on parenchymal enhancement however, have not been fully
investigated, especially in the light of low kV scanning and recent advances in
image reconstruction (e.g. iterative reconstruction) (19).

Scanner improvements aid CM volume optimization for each patient. Recent

technical advances allow for CT scans to be performed at tube voltages as
low as 70 kV, which decreases radiation dose significantly (20-22). In addition,

25

Vol_BM_productie.indd 25 02/02/2022 15:22:51

 Chapter 2

reducing the tube voltage increases CM enhancement, as the x-ray output

comes closer to the 33 keV k-edge of Iodine, which increases (liver) attenuation.
This facilitates reduction of CM volume whilst decreasing radiation dose (23,
24). Most previous research on the topic of TBW and liver attenuation has been
performed at a fixed tube voltage setting of 120 kV (6, 25, 26). Until now, most
thorough studies were performed in an Asian population, with a lower mean
TBW than an average European or American population (6, 25-28).

The aim of this study was to establish a possible benefit of individualized

CM injection over a fixed CM volume in liver imaging, when applying modern
scanner techniques in image acquisition (90 kV protocolling) and post-
processing (raw-data based iterative reconstruction methods) in a (heavier)
European population.

Materials and Methods

Ethics

The local ethical committee and institutional review board provided a waiver
of written informed consent for the study design, as the data was analysed
anonymously in accordance with the Institutional Review Board guidelines
(METC, ref. 16-4-161).

Study population

All abdominal CT scans in portal venous phase or in combination with a thoracic

CTA in patients ≥ 18 years of age, were eligible for inclusion. Patients were
excluded in case of hemodynamic instability and general exclusion criteria for
contrast enhanced CT were applied (e.g. pregnancy, renal insufficiency [eGFR,
30 ml/min/1.73 m2], iodine allergy). Technicians asked the patients’ weight
prior to the scan. In case of doubt, a weighing scale was available. Patients
were excluded when the inserted intravenous catheter was not capable of
reaching the necessary flow rate for the individualized CM injection, or when
injection data was not complete (n = 17). Two patients were excluded because
of CM extravasation. In total, 199 consecutive patients were enrolled between
November 2017 and May 2018.

26

Vol_BM_productie.indd 26 02/02/2022 15:22:51

 Body Weight–Adapted CM Application in CT Imaging

Imaging protocol

All scans were performed on a 3rd-generation dual-source CT (DSCT) scanner

(Somatom Force, Siemens Healthineers, Forchheim, Germany). Scan range was
set from approximately 2 cm cranial of the diaphragm to the pubic symphysis.
Scan parameters were as follows: tube voltage was 90 kV, 192 x 0.6 mm slice
collimation, gantry rotation time of 0.5 s and a quality reference tube current
of 295 mAsref (CareDose 4DTM, Siemens). Image reconstruction was performed
in the axial, coronal and sagittal plane with 3 mm slice thickness and 2 mm
increment using a Br40 kernel (Advanced Modelled Iterative Reconstruction 2
[ADMIRE], strength 2).

Contrast media injection protocol

All patients received pre-warmed CM (37 °C [99 °F]) (Ultravist®, Iopromide 300
mg/ml; Bayer Healthcare, Berlin, Germany). All scans were performed in the
portal venous phase with a fixed scan delay of 70 s after CM administration, or
approximately 35 s after the arterial phase of the thorax. For the latter, delay
was calculated by means of bolus tracking, whereas the abdominal scan was
performed after an average of 70 s after the start of the CM injection. CM was
administered using a programmable dual-head CT power injector (Stellant®;
Bayer) and injected through an 18, 20 or 22 Gauge needle, or through a central
line.

Group 1 received a standard fixed CM volume of 110 ml (TIL: 33 g I), with a flow
rate of 3.5 ml/s (IDR: 1.05 g I/s) followed by a saline flush of 40 ml at the same
flow rate.

Group 2 received a body weight adapted CM injection protocol as determined

by contrast injection software (P3T TM), which consisted of a CM phase followed
by a saline flush of 40 ml at the same flow rate (maximum flow rate: 6 ml/s). The
individually tailored CM injection software calculates the TIL and flow rate for
each patient, depending on body weight. The dosing factor was 0.4 g I/kg and
injection time was 30 seconds for all patients (14, 15).

27

Vol_BM_productie.indd 27 02/02/2022 15:22:51

 Chapter 2

A dedicated data acquisition program (CertegaTM Informatics Solution; Bayer)

continuously monitored and collected all injection parameters (e.g. total amount
of CM [ml] and peak flow rate [ml/s]).

Figure 1. ROIs were drawn in segment 2, 5 and 8 of the liver (when available). The white
circle indicates the ROI drawn to determine HU in a paraspinal muscle to determine
CNR. (ROI, region of interest; CNR, contrast-to-noise ratio)

Objective image quality

Image quality (IQ) was evaluated by measuring the attenuation (HU) in the
liver parenchyma, signal-to-noise (SNR) and contrast-to-noise (CNR) ratio.
One experienced researcher (B.M.) measured attenuation values by manually
delineating regions of interest (ROIs) within the liver parenchyma. Segments 2, 5
and 8, according to the Couinaud distribution, were used where possible (figure
1) (29). In case liver surgery was performed, the adjacent segment was chosen.
An ROI was drawn in each segment (≥ 1 cm2) without involvement of bordering
vascular structures. SNR was calculated by dividing the attenuation of the liver
parenchyma by the corresponding standard deviation (SD) of the attenuation
(30-34). The attenuation of the left erector spinae muscle was measured at
the level of the liver, in order to calculate CNR using the following established
formula: liver segment attenuation minus intramuscular attenuation, divided
by the SD of the intramuscular attenuation (16, 31-36).

Subjective image quality

Two radiologists (C.M. and B.M.), respectively with 8 and 3 years of experience
in abdominal radiology, evaluated the subjective IQ in consensus while blinded

28

Vol_BM_productie.indd 28 02/02/2022 15:22:51

 Body Weight–Adapted CM Application in CT Imaging

to the injection protocol. Subjective IQ was assessed by using a 5-point Likert

scale [1 = excellent; 2 = good; 3 = moderate; 4 = poor; 5 = very poor] (37).

Statistical Analysis

The Kolmogorov-Smirnov test was used to check for normal distribution.

Continuous variables were reported as mean ± SD for normally distributed
variables. Differences between groups were analysed with the unpaired
samples T-test or a one-way analysis of variance (ANOVA) with a Tukey test
for post hoc comparison, depending on the number of groups. For non- 2
normally distributed variables, the Mann-Whitney U or Kruskal-Wallis test was
performed. Categorical variables were reported as the number of cases and the
percentages per group, the X 2 test was used to calculate differences between
these groups. Statistic software (SPSS, version 24.0, IBM Corp., New York, NY)
was used for the data analysis. All p-values were two-sided and a p-value below
0.05 was considered statistically significant.

Results

Baseline characteristics

The baseline characteristics of the study population are depicted in table 1 for
both groups. No significant differences in baseline characteristics were found
between groups.

Table 1. Baseline characteristics. No significant differences were found between groups.

(BMI, body mass index; Abd/Th - Abd, Abdominal scan / Abdominal scan with a thoracic CT)

Patient Characteristics Group 1 Group 2 p-value

(N = 100) (N = 99)
Excluded patients 4 15
Age (years) 64.2 ± 16.1 64.5 ± 14.5 0.841
Sex (% male) 52 (52 %) 53 (54 %) 0.828
Body weight (kg) 77.1 ± 15.5 77.9 ± 15.9 0.713
Height (m) 1.70 ± 0.09 1.72 ± 0.09 0.188
BMI (kg m -2) 26.5 ± 4.3 26.3 ± 4.7 0.716

29

Vol_BM_productie.indd 29 02/02/2022 15:22:51

 Chapter 2

Table 1. Continued

Patient Characteristics Group 1 Group 2 p-value

(N = 100) (N = 99)
Scan indication
Oncology 79 (79 %) 81 (82 %) 0.658
Infectious 7 (7 %) 4 (4 %)
Other 14 (14 %) 14 (14 %)
Scan protocol (Abd/Th - Abd) 52 % / 48 % 55 % / 46 % 0.719
Needle Size
18 Gauge 4 (4 %) 6 (6 %) 0.828
20 Gauge 86 (86 %) 86 (87 %)
22 Gauge 1 (1 %) 1 (1 %)
Missing data 9 (9 %) 6 (6 %)

Injection parameters

Table 2 depicts the injection parameters per group and per weight category,
as all patients were divided into three weight categories; ≤ 70 kg, 71 - 85 kg
and ≥ 86 kg.

Table 2. Injection parameters. (CM, contrast media; IDR, iodine delivery rate; TIL, total
iodine load)

N CM volume TIL (g) Flow rate IDR (g Grams of

(ml) ± SD ± SD (ml/s) ± SD I/s) ± SD Iodine per kg
Group 1 ≤ 70 kg 36 110 33 3.5 1.05 0.55 ± 0.05
71 - 85 kg 36 110 33 3.5 1.05 0.42 ± 0.02
≥ 86 kg 28 110 33 3.5 1.05 0.35 ± 0.03
Group 2 ≤ 70 kg 35 82.4 ± 9.2 24.7 ± 2.7 ± 0.3 0.82 ± 0.4
2.8 0.1
71 - 85 kg 35 104.3 ± 5.3 31.3 ± 3.4 ± 0.2 1.03 ± 0.4
1.6 0.1
≥ 86 kg 29 130.0 ± 12.6 38.9 ± 4.2 ± 0.4 1.28 ± 0.4
3.4 0.1
p-value < 0.01 < 0.01 0.356 0.448 < 0.01
Group 1 and 2

30

Vol_BM_productie.indd 30 02/02/2022 15:22:51

 Body Weight–Adapted CM Application in CT Imaging

Figure 2 sets out the CM volume against weight for both groups. The mean
CM volume for group 2 was 104.1 ± 21.2 ml (range: 60.3 - 165.3 ml), which was
significantly lower than the CM volume in group 1 (110 ml) (p < 0.01). A CM
volume below 110 ml was used in 65.7% of the patients in group 2.

Figure 2. CM volume set out to weight for each group. Group 1 received a fixed CM
volume of 110 ml. Group 2 received a CM volume based on total body weight. (CM,
contrast media)

Table 3. Mean effective mAs, CTDIvol (mGy) and DLP (mGy*cm) shown per group and weight
category. Values increase with an increasing weight and no significant differences were found
between groups. (CTDIvol, CT dose index vol; DLP, dose length product)

Mean Effective Mean CTDIvol Mean DLP

mAs ± SD (mGy) ± SD (mGy*cm) ± SD
Group 1 ≤ 70 kg 161.6 ± 37.3 4.7 ± 1.1 217.2 ± 58.5
71 - 85 kg 222.4 ± 78.1 6.3 ± 2.4 311.7 ± 105.4
≥ 86 kg 239.5 ± 65.0 6.9 ± 1.9 351.2 ± 105.5
Mean 205.3 ± 70.2 5.9 ± 2.1 288.7 ± 106.4
Group 2 ≤ 70 kg 158.2 ± 42.1 4.6 ± 1.2 208.9 ± 50.4
71 - 85 kg 203.7 ± 39.9 5.9 ± 1.2 285.3 ± 59.7
≥ 86 kg 260.6 ± 102.7 7.5 ± 3.0 379.5 ± 145.5
Mean 204.3 ± 76.7 5.9 ± 2.2 285.9 ± 113.5
Group 1 vs p-value 0.969 0.799 0.950
Group 2

31

Vol_BM_productie.indd 31 02/02/2022 15:22:51

 Chapter 2

Radiation dose

Mean effective mAs (mAseff ), CT dose index vol (CTDIvol) and dose length product
(DLP) for group 1 were 205.3 ± 70.2 mAseff, 5.9 ± 2.1 mGy and 288.7 ± 106.4
mGy*cm, respectively. In group 2, mean values were 204.3 ± 76.7 mAseff, 5.9
± 2.2 mGy and 285.9 ± 113.5 mGy*cm. No significant differences were found
between groups (table 3).

Table 4. Attenuation value (HU), signal-to-noise ratio (SNR) and contrast-to-noise ratio
(CNR) for each group, shown per weight category. No significant differences in HU, SNR or
CNR were found between the two groups. Although for group 1, the attenuation differed
significantly between certain weight groups as mentioned below. (HU, Hounsfield Units;
SNR, signal-to-noise ratio; CNR, contrast-to-noise ratio).

Mean HU ± SD Mean SNR ± SD Mean CNR ± SD

Group 1 127.8 ± 23.7 8.5 ± 2.5 5.6 ± 2.9
Group 2 126.0 ± 17.4 8.2 ± 1.6 5.4 ± 2.1
p-value 0.536 0.369 0.518
Group 1 ≤ 70 kg 139.9 ± 21.4 10.4 ± 2.1 6.6 ± 2.7
71 - 85 kg 124.6 ± 24.8 8.0 ± 2.1 5.5 ± 3.3
≥ 86 kg 116.2 ± 17.8 6.6 ± 1.4 4.2 ± 1.8
p-value < 0.01 1
< 0.01 2
< 0.013
Group 2 ≤ 70 kg 126.5 ± 15.8 9.2 ± 1.2 6.0 ± 1.7
71 - 85 kg 128.2 ± 15.3 8.3 ± 1.1 5.8 ± 2.1
≥ 86 kg 122.7 ± 21.2 6.9 ± 1.7 4.3 ± 2.0
p-value 0.450 < 0.01 2
< 0.014
1
Post hoc comparison showed a significant difference between weight category ≤ 70 kg and
71 - 85 kg; and ≤ 70 kg and ≥ 86 kg.
2
Post hoc comparison showed a significant difference between all three weight categories.
3
Post hoc comparison showed a significant difference between weight category ≤ 70 kg
and ≥ 86 kg.
4
Post hoc comparison showed a significant difference between weight category ≤ 70 kg and
≥ 86 kg; and 71 - 85 kg and ≥ 86 kg.

Objective image quality

For group 1 mean attenuation values of the liver parenchyma for each weight
category (≤ 70 kg; 71 - 85 kg; ≥ 86 kg) were 139.9 ± 21.4 HU; 124.6 ± 24.8 HU
and 116.2 ± 17.8 HU. A significant difference in attenuation was found between

32

Vol_BM_productie.indd 32 02/02/2022 15:22:51

 Body Weight–Adapted CM Application in CT Imaging

the lowest and the middleweight category and between the lowest and highest
weight group. In contrast, group 2 attenuation values were comparable and not
significantly different between the three weight groups; 126.5 ± 15.8 HU; 128.2
± 15.3 HU and 122.7 ± 21.2 HU, respectively (p = 0.450, table 4 and figure 3).
The mean SNR and CNR were not statistically different between group 1 and
group 2 (p = 0.369 and 0.518, respectively) (table 4). The mean SNR for group 1
and 2 was 8.5 ± 2.5 (range: 1.9 - 14.5) and 8.2 ± 1.6 (range: 3.5 - 11.7). For CNR,
mean values were: 5.6 ± 2.9 (range: -5.4 - 16.8) and 5.4 ± 2.1 (range: 0.7 - 11.9)
for group 1 and 2.
2
Figure 3. Attenuation of the liver parenchyma in segment 2, 5, 8, according to the
Couinaud distribution (29). When liver surgery was performed, the adjacent segment
was chosen. Attenuation is set out per weight category for both group 1 and group 2.

Subjective Image quality

The subjective IQ was diagnostic in all scans, ranging from average to excellent
with no significant difference between groups (p = 0.213) (table 5). No significant
differences in subjective IQ between the weight categories were found in both
group 1 and group 2 (p = 0.076 and 0.358, respectively).

Table 5. Subjective IQ rated on a 5-point Likert scale for both groups. (IQ, image quality)

Excellent Good Average Poor Very Poor

Group 1 25 (25.0 %) 63 (63.0 %) 12 (12.0 %) 0 (0.0 %) 0 (0.0 %)
Group 2 15 (15.2 %) 72 (72.7 %) 12 (12.1 %) 0 (0.0 %) 0 (0.0 %)

33

Vol_BM_productie.indd 33 02/02/2022 15:22:52

 Chapter 2

Discussion

An individualized CM injection protocol tailored to TBW resulted in a more

homogeneous enhancement of the liver parenchyma in comparison to the
fixed CM volume injection protocol (figure 3). In the latter, a steady decline of
liver enhancement with increasing TBW was observed.

In the portal venous phase, CM volume and TIL are the most important factors
determining liver enhancement. Flow rate and IDR are less important, unlike
in CTA where those parameters are most influential (12, 23). Due to the nature
of our study design, CM volume and TIL are significantly different between
group 1 and 2, while values for flow rate and IDR are comparable between
both groups (p > 0.05) (table 2). The individualized protocol resulted in a CM
volume reduction for nearly two third of our patients while ensuring similar IQ.

The standard deviation is smaller in group 2, which is also an indication for a

more homogeneous attenuation of the liver parenchyma. Interestingly, the
highest weight category shows a similar standard deviation to group 1 and a
slightly lower overall attenuation than the other weight categories in group 2.
A potential explanation for this finding might be a higher percentage of people
with steatosis in the heavier population and therefore a greater spread in
attenuation in this category.

Some studies have already established the beneficial effect of using body size
parameters to individualize CM injection protocols in liver imaging. However,
most previous studies were performed in an Asian population and/or at a tube
voltage of 120 kV and/or with filtered back projection (6, 25, 26). Mean TBW in
the studies by Kondo et al. and Awai et al. ranged between 53.5 and 57.6 kg (6,
25, 26). Mean CM volume used in those studies was between 107 and 111 ml,
with a TIL between 32.1 and 33.3 g I (6, 26). Mean TBW in our population was
much higher than the mean body weight in the earlier mentioned Asian studies,
while in addition, we were able to use a lower mean CM volume. The use of a
standard lower tube voltage in combination with a body weight adapted CM
injection protocol and advanced iterative image reconstruction, resulted in
nearly a 5 % reduction of CM volume for group 2 compared to the Asian studies.

34

Vol_BM_productie.indd 34 02/02/2022 15:22:52

 Body Weight–Adapted CM Application in CT Imaging

Administrating too much CM in lighter patients can result in hyper attenuation

of the liver parenchyma and an unnecessarily high total injected CM volume.
Although this doesn’t necessarily lead to inadequate IQ, it is not preferable
for the patients. In the heavy patient population however, an insufficient CM
volume might result in a decreased detectability of liver lesions.

Recent literature does not describe a clear cut-off value for diagnostic IQ. Mean
SNR values range from 4.3 ± 0.6 to 17.9 ± 1.9 and mean CNR ranges between
5.2 ± 2.7 and 6.8 ± 3.0 in recent studies using iterative reconstructions (4, 27,
28, 32, 33, 38, 39). These values show a high degree of divergence and are not 2
comparable between studies, because different scanners, scan techniques
and CM injection protocols are used. However, in this study SNR and CNR were
not significantly different between both groups and consistent with previous
published data. Previous literature states the sole use of parameters such
as CNR and SNR might not be a correct representation of the IQ (40, 41). For
example, the CNR only depends on contrast and noise. Factors such as the
size of a lesion, its shape and the distribution of the CM attenuation within the
lesion are not taken in to account. This is considered a shortcoming in currently
used methods for determining objective IQ in CT imaging.

Currently, abdominal CT scans in daily clinical routine are performed at lower

tube voltages than the former clinical standard of 120 kV. Reducing tube voltage
most importantly results in a radiation dose reduction, but also provides the
possibility for CM volume reduction. Diagnostic accuracy, however, should
be prioritised over radiation dose and CM volume in liver lesion detection.
Maximal reduction of radiation and CM volume are of questionable value if
the radiologist can no longer differentiate between presence or absence of
liver lesions. Higher tube potentials, fuelled by the development of modern
scanners, in combination with tube current modulation software ensure that
the tube current can be increased to a great extent, guaranteeing a constant
IQ (20). However, no research has been performed to prove that this tube
voltage and CM volume reduction results in the same IQ and lesion detection
potential as the ground rules set out by Heiken et al. (5). Future research should
be tailored towards optimization of both radiation dose and CM volume while
maintaining diagnostic IQ.

35

Vol_BM_productie.indd 35 02/02/2022 15:22:52

 Chapter 2

Limitations

This study has several limitations. First, this is a single-centre study,

investigating a limited number of patients. In our opinion however, the baseline
characteristics are a good reflection of the European population. Secondly,
patients reported their own weight and only in case of doubt, a weighing
scale was used. Therefore, some discrepancy in patients’ weight could have
occurred. However, this is a straightforward approach which is comparable to
the clinical setting as well. Next, lean body weight has proven to be useful in
the Asian population, it would be interesting to investigate this parameter in
future studies and compare it to TBW. Fourth, liver diseases (e.g. steatosis and
cirrhosis) and other parameters, such as cardiac function, most likely influence
liver attenuation to a certain degree. These patients were not excluded here,
but assumed to be randomly assigned to both groups. Therefore, it could
be interesting to have a closer look into this patient subpopulation, e.g. by
analysing delta HU in attenuation between unenhanced and a portal venous
phase CT. This delta HU could provide a more constant parameter to determine
liver enhancement, compared to HU in portal venous phase solely.

Conclusion

Usage of a body weight tailored CM injection protocol results in more

homogeneous liver enhancement at lower tube voltage (e.g. 90 kV) in
comparison to a fixed CM injection protocol, while CM volume can be reduced
in a large percentage of the population.

36

Vol_BM_productie.indd 36 02/02/2022 15:22:52

 Body Weight–Adapted CM Application in CT Imaging

References

1. Schulz A, Viktil E, Godt JC, et al. Diagnostic performance of CT, MRI and PET/CT in
patients with suspected colorectal liver metastases: the superiority of MRI. Acta Radiol.
2015:1040-8.
2. Chou R, Cuevas C, Fu R, et al. Imaging techniques for the diagnosis of hepatocellular
carcinoma: a systematic review and meta-analysis. Ann Intern Med. 2015;162(10):697-711.
3. Robinson E, Babb J, Chandarana H, Macari M. Dual source dual energy MDCT:
comparison of 80 kVp and weighted average 120 kVp data for conspicuity of hypo-
vascular liver metastases. Invest Radiol. 2010;45(7):413-8.
4. Chen CY, Hsu JS, Jaw TS, et al. Lowering radiation dose during dedicated colorectal cancer 2
MDCT: comparison of low tube voltage and sinogram-affirmed iterative reconstruction
at 80 kVp versus blended dual-energy images in a population of patients with low body
mass index. Abdom Imaging. 2015;40(7):2867-76.
5. Heiken JP, Brink JA, McClennan BL, et al. Dynamic incremental CT: effect of volume
and concentration of contrast material and patient weight on hepatic enhancement.
Radiology. 1995;195(2):353-7.
6. Kondo H, Kanematsu M, Goshima S, et al. Body size indexes for optimizing iodine dose
for aortic and hepatic enhancement at multidetector CT: comparison of total body
weight, lean body weight, and blood volume. Radiology. 2010;254(1):163-9.
7. Bae KT, Shah AJ, Shang SS, et al. Aortic and hepatic contrast enhancement with
abdominal 64-MDCT in pediatric patients: effect of body weight and iodine dose. AJR
Am J Roentgenol. 2008;191(5):1589-94.
8. Walkey MM. Dynamic hepatic CT: how many years will it take ‘til we learn? Radiology.
1991;181(1):17-8.
9. Kondo H, Kanematsu M, Goshima S, et al. Aortic and hepatic enhancement at
multidetector CT: evaluation of optimal iodine dose determined by lean body weight.
Eur J Radiol. 2011;80(3):e273-7.
10. Bae KT, Heiken JP, Brink JA. Aortic and hepatic contrast medium enhancement at CT. Part
II. Effect of reduced cardiac output in a porcine model. Radiology. 1998;207(3):657-62.
11. Behrendt FF, Mahnken AH, Keil S, et al. Contrast enhancement in multidetector-row
computed tomography (MDCT) of the abdomen: intraindividual comparison of contrast
media containing 300 mg versus 370 mg iodine per ml. Eur Radiol. 2008;18(6):1199-205.
12. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
13. Seo N, Chung YE, Lim JS, et al. Bowel aangioedema associated with iodinated contrast
media: incidence and predisposing factors. Invest Radiol. 2017;52(9):514-21.
14. Seifarth H, Puesken M, Kalafut JF, et al. Introduction of an individually optimized
protocol for the injection of contrast medium for coronary CT angiography. Eur Radiol.
2009;19(10):2373-82.

37

Vol_BM_productie.indd 37 02/02/2022 15:22:52

 Chapter 2

15. Mihl C, Kok M, Altintas S, et al. Evaluation of individually body weight adapted contrast
media injection in coronary CT-angiography. Eur J Radiol. 2016;85(4):830-6.
16. Hendriks BM, Kok M, Mihl C, et al. Individually tailored contrast enhancement in CT
pulmonary angiography. Br J Radiol. 2016;89(1061):20150850.
17. Kok M, Mihl C, Hendriks BM, et al. Optimizing contrast media application in coronary
CT angiography at lower tube voltage: evaluation in a circulation phantom and sixty
patients. Eur J Radiol. 2016;85(6):1068-74.
18. Hendriks BMF, Eijsvoogel NG, Kok M, et al. Optimizing pulmonary embolism computed
tomography in the age of individualized medicine: a prospective clinical study. Invest
Radiol. 2018;53(5):306-12.
19. Morsbach F, Desbiolles L, Raupach R, et al. Noise texture deviation: a measure for
quantifying artifacts in computed tomography images with iterative reconstructions.
Invest Radiol. 2017;52(2):87-94.
20. Lell MM, Wildberger JE, Alkadhi H, et al. Evolution in computed tomography: the battle
for speed and dose. Invest Radiol. 2015;50(9):629-44.
21. Saltybaeva N, Schmidt B, Wimmer A, et al. Precise and automatic patient positioning in
computed tomography: avatar modeling of the patient surface using a 3-dimensional
camera. Invest Radiol. 2018.
22. Attenberger UI, Morelli J, Budjan J, et al. Fifty years of technological innovation: potential
and limitations of current technologies in abdominal magnetic resonance imaging and
computed tomography. Invest Radiol. 2015;50(9):584-93.
23. Fleischmann U, Pietsch H, Korporaal JG, et al. Impact of contrast media concentration
on low-kilovolt computed tomography angiography: a systematic preclinical approach.
Invest Radiol. 2018;53(5):264-70.
24. Schmidt BT, Hupfer M, Saltybaeva N, et al. Dose optimization for computed tomography
localizer radiographs for low-dose lung computed tomography examinations. Invest
Radiol. 2017;52(2):81-6.
25. Awai K, Kanematsu M, Kim T, et al. The optimal body size index with which to determine
iodine dose for hepatic dynamic CT: a prospective multicenter study. Radiology.
2016;278(3):773-81.
26. Kondo H, Kanematsu M, Goshima S, et al. Body size indices to determine iodine mass
with contrast-enhanced multi-detector computed tomography of the upper abdomen:
does body surface area outperform total body weight or lean body weight? Eur Radiol.
2013;23(7):1855-61.
27. Goshima S, Kanematsu M, Noda Y, et al. Minimally required iodine dose for the
detection of hypervascular hepatocellular carcinoma on 80-kVp CT. AJR Am J Roentgenol.
2016;206(3):518-25.
28. Goshima S, Kanematsu M, Noda Y, et al. Determination of optimal intravenous contrast
agent iodine dose for the detection of liver metastasis at 80-kVp CT. Eur Radiol.
2014;24(8):1853-9.

38

Vol_BM_productie.indd 38 02/02/2022 15:22:52

 Body Weight–Adapted CM Application in CT Imaging

29. Germain T, Favelier S, Cercueil JP, et al. Liver segmentation: practical tips. Diagn Interv
Imaging. 2014;95(11):1003-16.
30. Shuman WP, Chan KT, Busey JM, et al. Standard and reduced radiation dose liver CT
images: adaptive statistical iterative reconstruction versus model-based iterative
reconstruction-comparison of findings and image quality. Radiology. 2014;273(3):793-
800.
31. Szucs-Farkas Z, Strautz T, Patak MA, et al. Is body weight the most appropriate criterion
to select patients eligible for low-dose pulmonary CT angiography? Analysis of objective
and subjective image quality at 80 kVp in 100 patients. Eur Radiol. 2009;19(8):1914-22.
32. Scholtz JE, Wichmann JL, Husers K, et al. Automated tube voltage adaptation in
combination with advanced modeled iterative reconstruction in thoracoabdominal
2
third-generation 192-slice dual-source computed tomography: effects on image quality
and radiation dose. Acad Radiol. 2015;22(9):1081-7.
33. Zhang X, Li S, Liu W, et al. Double-low protocol for hepatic dynamic CT scan: effect of low
tube voltage and low-dose iodine contrast agent on image quality. Medicine (Baltimore).
2016;95(26):e4004.
34. Song JS, Lee JM, Sohn JY, et al. Hybrid iterative reconstruction technique for liver CT scans
for image noise reduction and image quality improvement: evaluation of the optimal
iterative reconstruction strengths. Radiol Med. 2015;120(3):259-67.
35. Tawfik AM, Kerl JM, Bauer RW, et al. Dual-energy CT of head and neck cancer: average
weighting of low- and high-voltage acquisitions to improve lesion delineation and image
quality-initial clinical experience. Invest Radiol. 2012;47(5):306-11.
36. Pasquier H, Gardavaud F, Chiaradia M, et al. Iterative reconstructions in multiphasic CT
imaging of the liver: qualitative and task-based analyses of image quality. Clin Radiol.
2018;73(9):834 e9- e16.
37. Caruso D, De Santis D, Rivosecchi F, et al. Lean Body Weight-Tailored Iodinated
Contrast Injection in Obese Patient: Boer versus James Formula. Biomed Res Int.
2018;2018:8521893.
38. Nakamoto A, Kim T, Hori M, et al. Clinical evaluation of image quality and radiation
dose reduction in upper abdominal computed tomography using model-based iterative
reconstruction; comparison with filtered back projection and adaptive statistical
iterative reconstruction. Eur J Radiol. 2015;84(9):1715-23.
39. Kanematsu M, Kondo H, Miyoshi T, et al. Whole-body CT with high heat-capacity X-ray
tube and automated tube current modulation--effect of tube current limitation on
contrast enhancement, image quality and radiation dose. Eur J Radiol. 2015;84(5):877-83.
40. Vaishnav JY, Jung WC, Popescu LM, et al. Objective assessment of image quality and
dose reduction in CT iterative reconstruction. Med Phys. 2014;41(7):071904.
41. De Crop A, Smeets P, Van Hoof T, et al. Correlation of clinical and physical-technical
image quality in chest CT: a human cadaver study applied on iterative reconstruction.
BMC Med Imaging. 2015;15:32.

39

Vol_BM_productie.indd 39 02/02/2022 15:22:52

Vol_BM_productie.indd 40 02/02/2022 15:22:53
 CHAPTER
3
A solution for homogeneous
liver enhancement in computed
tomography

Results from the COMpLEx Trial

B. Martens, J.E. Wildberger, B.M.F. Hendriks, S.M.J. Van Kuijk, E.C. Nijssen,
N.H.G.M. Peters, J. De Vos – Geelen, C. Mihl

Published in:
Investigative Radiology 2020; 55(10):666-672

Vol_BM_productie.indd 41 02/02/2022 15:22:55

 Chapter 3

Abstract

Objectives: The aim of the study was to reach homogeneous enhancement of

the liver, irrespective of total body weight (TBW) or tube voltage. An easy to use
rule of thumb, the 10-to-10 rule, which pairs a 10 kV reduction in tube voltage
with a 10 % decrease in contrast media (CM) dose, was evaluated.

Materials and Methods: Two hundred fifty-six patients scheduled for an

abdominal CT in portal venous phase were randomly allocated to one of four
groups. In group 1 (n = 64) a tube voltage of 120 kV and a TBW-adapted CM
injection protocol was used: 0.521 g I/kg. In group 2 (n = 63), tube voltage was
90 kV and the TBW-adapted CM dosing factor remained 0.521 g I/kg. In group
3 (n = 63), tube voltage was reduced by 20 kV and CM dosing factor by 20 %
compared to group 1, in line with the 10-to-10 rule (100 kV; 0.417 g I/kg). In
group 4 (n = 66): tube voltage was decreased by 30 kV paired with a 30 %
decrease in CM dosing factor compared to group 1, in line with the 10-to-10
rule (90 kV; 0.365 g I/kg). Objective image quality was evaluated by measuring
attenuation in Hounsfield units (HU), signal-to-noise ratio (SNR) and contrast-
to-noise ratio (CNR) in the liver. Overall subjective image quality was assessed
by two experienced readers by using a 5-point Likert scale. Two-sided p-values
below 0.05 were considered significant.

Results: Mean attenuation values in groups 1, 3 and 4 were comparable (118.2 ±

10.0; 117.6 ± 13.9; 117.3 ± 21.6 HU respectively), whereas attenuation in group
2 (141.0 ± 18.2 HU) was significantly higher than all other groups (p < 0.01). No
significant difference in attenuation was found between weight categories ≤
80 kg and > 80 kg within the four groups (p ³ 0.371). No significant differences
in subjective image quality were found (p = 0.180).

Conclusions: The proposed 10-to-10 rule is an easily reproducible method

resulting in similar enhancement in portal venous CT of the liver throughout
the patient population, irrespective of TBW or tube voltage.

42

Vol_BM_productie.indd 42 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

Keywords

Multidetector Computed Tomography; Diagnostic Imaging; Liver; Radiation

Dosage; Contrast Media

Introduction

Contrast media (CM) are used in Computed Tomography (CT) scans to enhance
vascular structures and organ parenchyma. The visibility of liver lesions depends
mainly on image noise and the ratio between size and difference in attenuation
of the lesion compared to the parenchyma (1). Comparing the unenhanced
parenchyma to that after CM administration (in the same patient), Heiken et 3
al. (1995) found that an attenuation difference (∆) of at least 50 Hounsfield
Units (HU) is necessary to safely detect liver lesions (2). A dosing factor of 0.521
grams of iodine per kg (g I/kg) was proposed to reach the required ∆ 50 HU
at a given tube voltage of 120 kV (2). By taking the HU of the unenhanced liver
into account, a correction can be performed for any liver disorder that might
affect background attenuation of the liver.

Parenchymal enhancement depends on scan (e.g. CT scanner, tube voltage), CM

(e.g. volume, concentration, flow rate, temperature) and patient characteristics.
Relevant patient related parameters include weight, height, venous access,
cardiac output, age, gender, breath-hold, renal function and comorbidity (3).
Previous research showed that individualized CM injection protocols, where the
CM bolus is adapted to patient total body weight (TBW), lean body weight (LBW),
or body surface area (BSA), yields better results (1, 4-8). A recent feasibility
study demonstrated that a TBW adapted CM injection protocol resulted in more
homogeneous liver enhancement compared to fixed iodine load (9).

Recent technological developments in X-ray tube technology permit lower tube

voltages whilst maintaining satisfactory image quality, which subsequently leads
to lower radiation doses (10, 11). Reducing tube voltage increases attenuation
of iodine, by approaching the 33 keV k-edge of iodine. This enables both a
reduction of the radiation dose and CM volume (12). This phenomenon,
where changing tube voltage influences iodine attenuation, might result in
clinical controversies. For example, in imaging of renal masses attenuation

43

Vol_BM_productie.indd 43 02/02/2022 15:22:55

 Chapter 3

may indicate whether a lesion is more likely benign or malignant (13, 14).
When patients are scanned with variable tube voltages iodine attenuation is
affected, consequently conclusions cannot be derived from the magnitude
of the attenuation. Therefore, it is important to find a method by which the
attenuation pattern of parenchymal structures remains robust irrespective of
the tube voltage or patient TBW.

In recent literature the importance of individualized CM injection protocols

reducing CM volume at a lower tube potential has been stressed in vascular
studies (15-17). To the best of our knowledge this has not been investigated in
abdominal imaging. This study tested the following hypothesis: a 10 % reduction
in CM dosing factor per 10 kV should yield homogeneous enhancement of the
liver in portal venous CT, irrespective of TBW and at variable tube voltages
(10-to-10 rule) (18, 19).

The aim of the present study was to investigate whether adapting a TBW-
based dosing factor to the tube voltage used results in homogeneous liver
enhancement between patients.

Materials and Methods

Ethics

This double-blind randomized controlled trial was approved by the local ethics
committee as well as by the institutional review board and is registered on
ClinicalTrials.gov (NCT02462044). Written informed consent for inclusion in the
clinical trial was obtained.

Study population

Patients were enrolled between December 2018 and June 2019 at Maastricht
University Medical Center. Patients scheduled for an abdominal CT in the
portal venous phase were eligible for inclusion. Possible scan indications were
oncology, infection, and screening after incidental findings on ultrasound,
weight loss, or abdominal pain. Exclusion criteria were age below 18 years, TBW
> 115 kg (because of practical considerations: a CM syringe contains 200 ml),

44

Vol_BM_productie.indd 44 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

hemodynamic instability, and general contraindications for contrast-enhanced

CT (e.g. pregnancy, renal insufficiency [estimated glomerular filtration rate of <
30 mL/min per 1.73 m2] and iodine allergy). Scanning additional to the portal
venous phase was not a reason for exclusion (other phases: e.g. arterial phase,
late phase; other organ region: e.g. combination with thoracic scanning). Patient
body weight was measured on calibrated scales in the scanner room and the
patient height was asked prior to the CT scan. BMI was calculated by dividing
body weight (in kg) by height (in m) squared. Repeat inclusion was not expected
to influence study outcome and was therefore allowed.

Patients were prospectively included into one of four groups. A computer

random number generator prepared the randomization schedule in a 1:1:1:1 3
manner (i.e. balanced randomization). Stratification was performed, based
on age (< 60 and ≥ 60 years) and weight (< 75 and ≥ 75 kg). Variable block
randomization distributed patients equally over the groups.

Scan and contrast media protocol

All scans were performed on a third-generation dual-source CT scanner

(Somatom Force; Siemens Healthineers, Forchheim, Germany). Automated
tube current modulation (ATCM) was used (CareDose 4D; Siemens), while tube
voltage was set. A 3 mm slice was scanned at the level of the main portal
vein prior to CM administration to establish the baseline attenuation of the
unenhanced liver as mentioned in the introduction. Parameters were similar
to the subsequent contrast-enhanced scan: tube voltage 120, 100 or 90 kV
(depending on the allocated group); slice collimation 192 x 0.6 mm; gantry
rotation time 0.5 seconds; quality reference kV and mAs set to respectively 120
kVref and 150 mAsref. The abdominal scan range was set from approximately 2
cm above the diaphragm to the pubic symphysis.

Prewarmed CM (37°C [99°F]) was used at a concentration of 300 mg/mL

(Iopromide; Bayer Healthcare, Berlin, Germany). CM was injected with a
programmable dual-head CT power injector (Stellant, Bayer) through an 18,
20 or 22 gauge needle. Group 1 received the protocol considered the golden
standard: 120 kV and 0.521 g I/kg (2). Group 2 received an adapted protocol
with CM dosing factor identical to group 1 (e.g. 0.521g I/kg), but tube voltage

45

Vol_BM_productie.indd 45 02/02/2022 15:22:55

 Chapter 3

reduced to 90 kV. In group 3 tube voltage was set at 100 kV and the dosing
factor was reduced by 20 % in accordance with the 10-to-10 rule (e.g. 0.417 g
I/kg). Group 4 received a 90 kV scan protocol with a 30 % reduction in dosing
factor compared to group 1 in accordance with the 10-to-10 rule: 0.365 g I/kg
(figure 1). CM injection duration was 30 seconds in all patients, as determined
by dedicated CM injection software (P3T; Bayer Healthcare, Berlin, Germany),
and therefore flow rate (in ml/s) was dependent on the weight of the patient
and the allocated group (9). The scan in the portal venous phase was performed
70 seconds after start of the CM injection in all patients. CM volume (in ml),
total iodine load (TIL, g I), flow rate and iodine delivery rate (IDR, in g I/s) were
monitored and collected with a dedicated data acquisition program (CertegaTM
Informatics Solution; Bayer).

Figure 1. Patients were randomly assigned to one of four groups. An unenhanced slice
at the level of the portal vein was scanned before contrast media injection.

Dose-related parameters (e.g. CT dose index [CTDIvol, in mGy] and dose length
product [DLP, in mGy*cm]) were recorded and collected from the dose sheet
available at the PACS workstation (IMPAX version 6.6.1.5003, AGFA HealthCare
N.V., Mortsel, Belgium). As mentioned above, all patients scheduled for an
abdominal CT in portal venous phase were eligible for inclusion. Therefore, an

46

Vol_BM_productie.indd 46 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

additional thoracic scan or other scan phases of the liver were not reasons for
exclusion. As a result, three different dose protocols were possible: abdominal
scan in portal venous phase, abdominal scan in portal venous phase with
a separated arterial thoracic CT, or a thoracic and abdominal scan in portal
venous phase. Only the CTDIvol and DLP of the abdominal scan in portal venous
phase were collected from the dose sheet. In cases where the thorax and
abdomen where scanned together in portal venous phase, the corresponding
CTDIvol and DLP were collected.

Image reconstruction was performed with 3 mm slice thickness, with

overlapping increment of 2 mm, in an axial, coronal and sagittal plane with a
soft tissue kernel (Br40; Siemens; Advanced Modelled Iterative Reconstruction, 3
strength 2 - 3).

Data processing

The objective image quality was evaluated by measuring attenuation (HU) in

three different liver segments on both the unenhanced and contrast-enhanced
portal venous phase scans, where possible in segments 2, 5 and 8, according
to the Couinaud classification (20). If not possible (e.g. previous surgery, large
lesions) an adjacent location close to the respective segment was chosen.
A region of interest (ROI) was drawn in each liver segment (area: ≥ 1 cm2),
choosing the largest possible ROI area not containing large blood vessels,
bile ducts or liver lesions. Dividing the HU of each segment by its standard
deviation (SD) resulted in the signal-to-noise ratio (SNR) (21-25). The mean
of the measurements in segments 2, 5 and 8 is reported as the SNR. The
HU and SD of the left paraspinal muscle at the level of the liver was used to
calculate the contrast-to-noise ratio (CNR) as follows: the attenuation of each
liver segment minus the attenuation of the left paraspinal muscle, divided by
the SD of the attenuation of the paraspinal muscle (9, 22-27). The mean of three
CNR measurements is reported.

Two abdominal radiologists (B.M. and C.M.) with respectively 4- and 9-years’
experience in abdominal CT, rated the scans in portal venous phase in
consensus while being blinded to the protocol. The radiologists were allowed to

47

Vol_BM_productie.indd 47 02/02/2022 15:22:55

 Chapter 3

adjust window-level settings. Overall image quality was rated on a 5-point Likert
scale: 1 = excellent; 2 = good; 3 = moderate; 4 = poor; 5 = very poor (9, 28).

Statistical Analysis

Continuous variables are presented as mean ± SD and categorical variables

as absolute numbers with percentages. In order to correct for the possible
confounders gender and iterative reconstruction (IR) strength an analysis of
covariance was performed, since all variables are continuous. Fifteen patients
(5.9 %) were reconstructed with IR strength 3 instead of 2. It was decided not
to exclude the scans reconstructed with IR 3 but to statistically correct for this
inconvenience instead, as the IR strength does not influence the attenuation of
the liver parenchyma, which was our primary outcome (29). This analysis was
used for both continuous and ordinal variables, because the steps within the
ordinal variables were deemed to be of comparable size. P-values are all 2-sided
and considered significant when below 0.05. Statistical software (SPSS, version
24.0; IBM Corp, New York, NY) was used for the data analysis.

Table 1. Baseline characteristics (BMI indicates body mass index).

Patient Characteristics Group 1 Group 2 Group 3 Group 4

(N = 64) (N = 63) (N = 63) (N = 66)
Excluded patients 4 5 4 4
Age (years) 64.0 ± 11.4 66.1 ± 12.6 65.6 ± 8.5 64.3 ± 9.9
Sex (% male) 73.3 % 53.4 % 40.7 % 59.7 %
Body Weight (kg) 79.5 ± 12.7 77.7 ± 14.0 78.5 ± 14.3 79.8 ± 14.8
Height (m) 1.75 ± 0.1 1.71 ± 0.1 1.71 ± 0.1 1.74 ± 0.1
BMI (kg m-2) 25.8 ± 3.3 26.5 ± 4.2 26.7 ± 4.3 26.5 ± 4.5
Scan indication
Oncology (%) 95.0 % 89.7 % 94.9 % 96.8 %
Other (%) 5.0 % 10.3 % 5.1 % 3.2 %
Needle Size
18 gauge (%) 58.3 % 48.3 % 52.5 % 46.8 %
20 gauge (%) 31.7 % 37.9 % 39.0 % 40.3 %
22 gauge (%) 0.0 % 0.0 % 0.0 % 3.2 %
Missing data (%) 10.0 % 13.8 % 8.5 % 9.7 %

48

Vol_BM_productie.indd 48 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

Results

Two hundred fifty-six patients were randomly allocated to one of four groups
(group 1, n = 64; group 2, n = 63; group 3, n = 63; and group 4, n = 66) (table
1). Despite randomisation, we observed a difference in gender distribution
between groups (% male group 1 = 73.3; group 2 = 53.4; group 3 = 40.7; and
group 4 = 59.7). Fifteen patients were excluded: 12 for technical reasons; 2
because only the liver was imaged and therefore radiation doses where not
comparable; 1 because of CM extravasation.

Table 2. Data are presented as mean and standard deviation (SD). * Post hoc comparison
showed a significant difference between groups 1 and 3; groups 1 and 4; groups 2 and 3; 3
groups 2 and 4; and groups 3 and 4. Only groups 1 and 2 did not significantly differ. (CM
indicates contrast media; TIL, total iodine load; IDR, iodine delivery rate; CTDIvol, CT dose
index vol; DLP, dose length product).

Group 1 Group 2 Group 3 Group 4 p-value

(n = 60) (n = 58) (n = 59) (n = 62) for dif-
ference
between
groups
CM volume (ml) 138.0 ± 135.0 ± 109.1 ± 97.1 ± < 0.01
22.0* 24.3* 19.9* 18.0*
TIL (g) 41.4 ± 6.6* 40.5 ± 7.3* 32.7 ± 6.0* 29.1 ± 5.4* < 0.01
Flow rate (ml/s) 4.5 ± 0.7* 4.4 ± 0.8* 3.6 ± 0.7* 3.2 ± 0.6* < 0.01
IDR (g I/s) 1.4 ± 0.2 *
1.3 ± 0.2 *
1.1 ± 0.2 *
1.0 ± 0.2 *
< 0.01
PvP Number of 13 (21.7 %) 17 (29.3 %) 12 (20.3 %) 15 (24.2 %)
Abdomen patients (in %)
CTDIvol (mGy) 7.8 ± 1.0 7.8 ± 1.8 7.9 ± 2.2 6.6 ± 1.7 0.322
DLP (mGy*cm) 376.9 ± 339.0 ± 389.8 ± 303.2 ± 0.440
74.5 128.4 136.1 101.9
PvP Number of 18 (30.0 %) 14 (24.1 %) 20 (33.9 %) 22 (35.5 %)
Abdomen patients (in %)
+ Art.
thorax
CTDIvol (mGy) 7.4 ± 1.2 7.3 ± 2.4 6.5 ± 1.5 8.0 ± 3.2 0.308
DLP (mGy*cm) 349.7 ± 360.1 ± 310.1 ± 391.9 ± 0.459
63.7 137.8 67.7 155.7
PvP Number of 29 (48.3 %) 27 (46.6 %) 27 (45.8 %) 25 (40.3 %)
Thorax + patients (in %)
Abdomen
CTDIvol (mGy) 7.0 ± 1.6 6.2 ± 1.8 6.8 ± 2.2 7.0 ± 3.5 0.765

49

Vol_BM_productie.indd 49 02/02/2022 15:22:55

 Chapter 3

Table 2. Continued

Group 1 Group 2 Group 3 Group 4 p-value

(n = 60) (n = 58) (n = 59) (n = 62) for dif-
ference
between
groups
DLP (mGy*cm) 483.8 ± 386.2 ± 432.4 ± 472.2 ± 0.522
121.3 137.4 131.6 228.8
Mean CTDIvol (mGy) 7.3 ± 1.4 6.9 ± 2.0 6.9 ± 2.0 7.2 ± 3.1 0.405
DLP (mGy*cm) 420.4 ± 366.0 ± 382.3 ± 402.8 ± 0.178
114.9 134.1 125.5 189.2

Injection parameters and radiation dose

See table 2 for an overview of CM injection parameters. As a result of the study

design significant differences were found in CM volume, TIL, flow rate and IDR
with p-values < 0.01. Table 2 shows the scan protocols and the mean radiation
dose for each group. As expected with identical reference kV and mAs for
each group, no significant differences in volumetric CTDIvol or DLP were found
between groups, p = 0.405 and p = 0.178, respectively.

Table 3. Mean attenuation (HU), SNR and CNR between groups.

Group 1 Group 2 Group 3 Group 4 p-value

(n = 60) (n = 58) (n = 59) (n = 62)
Mean HU
Unenhanced 60.6 ± 7.2 56.0 ± 11.4 56.2 ± 10.3 53.7 ± 13.5 0.149
Mean HU PvP 118.2 ± 10.0 141.0 ± 18.2 117.6 ± 13.9 117.3 ± 21.6 < 0.011
Mean SNR PvP 9.3 ± 1.6 9.6 ± 1.9 8.8 ± 1.7 8.6 ± 2.1 < 0.012
Mean CNR PvP 5.8 ± 1.8 6.8 ± 2.2 5.4 ± 1.7 5.4 ± 2.7 < 0.013
1
Post hoc comparison showed a significant difference between groups 1 and 2 (p < 0.01);
groups 2 and 3 (p < 0.01); and groups 2 and 4 (p < 0.01).
2
Post hoc comparison showed a significant difference between groups 1 and 4 (p = 0.016);
groups 2 and 3 (p = 0.012); and groups 2 and 4 (p < 0.01).
3
Post hoc comparison showed a significant difference between groups 1 and 2 (p < 0.01);
groups 2 and 3 (p < 0.01); and groups 2 and 4 (p < 0.01).
(HU indicates Hounsfield Units; SD, standard deviation; PvP, portal venous phase; SNR
signal-to-noise ratio; and CNR contrast-to-noise ratio)

50

Vol_BM_productie.indd 50 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

Objective image quality

The mean HU in the portal venous phase was not significantly different between
groups 1, 3 and 4, whereas attenuation in group 2 was significantly higher
compared to all other three groups (table 3 and figure 2). Mean HU values
in the portal venous phase were 118.2 ± 10.0, 141.0 ± 18.2, 117.6 ± 13.9 and
117.3 ± 21.6 in groups 1, 2, 3 and 4 respectively. A significant difference in HU
was found between groups 1 and 2; between groups 2 and 3; and between
groups 2 and 4 (all p < 0.01). Mean body weight was approximately 80 kg in all
groups, and therefore patients were divided in two weight categories (≤ 80 kg
and > 80 kg), these were slightly different from the stratification factors used
for the randomization process (<75 and ≥ 75 kg). No significant difference in 3
attenuation in the portal venous phase between weight categories was found
within groups, with p-values 0.371, 0.925, 0862, and 0.557 for groups 1 through
4 respectively. Figure 2 depicts mean HU values in the portal venous phase,
per group and weight category. Mean HU values found for unenhanced slices
of the liver at the level of the main portal vein were not significantly different
between the four groups (p = 0.149).

Mean SNR was highest in groups 1 and 2 (9.3 ± 1.6 and 9.6 ± 1.9 respectively),
and significantly higher than the values in groups 3 and 4 (8.8 ± 1.7 and 8.6 ±
2.1 respectively, p < 0.01). CNR was significantly higher in group 2 (6.8 ± 2.2),
compared to groups 1, 3 and 4 (5.8 ± 1.8, 5.4 ± 1.7 and 5.4 ± 2.7 respectively,
p < 0.01) (table 3).

Subjective Image quality

Results of the subjective image quality evaluation are presented in table 4. No

significant differences were found between groups (p = 0.180). All scans were
regarded as diagnostic, none of the CT scans were rated of poor or very poor
image quality, and image quality was considered good or excellent in 93.7 %
of the scans.

51

Vol_BM_productie.indd 51 02/02/2022 15:22:55

 Chapter 3

Figure 2. Mean attenuation of the liver parenchyma in portal venous phase set out
per group and weight category.

Table 4. Subjective image quality scored in consensus.

Group 1 Group 2 Group 3 Group 4 p-value

(n = 60) (n = 58) (n = 59) (n = 62)
Excellent 18 (30.0 %) 9 (15.5 %) 16 (27.1 %) 11 (17.7 %)
Good 41 (68.3 %) 44 (75.9 %) 38 (64.4 %) 47 (75.8 %)
Moderate 1 (1.7 %) 5 (8.6 %) 5 (8.5 %) 4 (6.5 %) 0.180
Poor 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 0 (0.0 %)
Very Poor 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 0 (0.0 %)

52

Vol_BM_productie.indd 52 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

Discussion

This study showed that an individualized CM injection and scan protocol, where
a 10 kV reduction in tube voltage is paired with a 10 % reduction in dosing
factor, resulted in homogeneous enhancement of the liver throughout the
entire study population. By using this 10-to-10 rule and the CM dosing factor,
portal venous abdominal CT protocols can be easily individualized based on
tube voltage and patient TBW.

As hypothesized, the 10-to-10 rule results in robust enhancement of the liver

at variable tube voltages irrespective of TBW. This is illustrated in figure 3,
which shows two scans of a patient who was included twice in the study and 3
allocated to two different scan protocols resulting in similar enhancement of
the liver (first allocation to group 4: 90 kV and 0.365 g I/kg; second allocation
to group 3: 110 kV and 0.417 g I/kg).

Figure 3. A 57-year old man in the follow up for metastasized urothelial cell carcinoma,
included twice and randomized in two different groups. Images were both reconstructed
with kernel BR40 and iterative reconstruction (IR) strength 2. The circle indicates the mean
Hounsfield Units (HU) measured in three different liver segments (preferable in segment
2, 5 and 8, according to the Couinaud distribution [19]), with the mean standard deviation.

Mean HU values in the portal venous phase were not significantly different
between groups 1, 3 and 4, while attenuation was significantly higher in group
2 compared to the other three groups. In addition, when this rule is applied,

53

Vol_BM_productie.indd 53 02/02/2022 15:22:55

 Chapter 3

a 10 % CM dose reduction can be achieved with every 10 kV tube voltage

reduction (figure 2). Mean HU values for the unenhanced slice of the liver were
not significantly different between groups and we may conclude that possible
factors influencing attenuation of the unenhanced liver (e.g. steatosis and
cirrhosis) were not noticeably different between groups and will not unduly
influence attenuation in portal venous phase.

CNR was highest in group 2 and comparable between groups 1, 2 and 3 (table
3). The larger variation in SNR values can be explained by the study setup. A
higher tube voltage with comparable tube current results in less image noise,
while higher CM volumes result in a higher attenuation. Therefore, SNR is, as
expected, highest in group 2. In group 4, the lowest tube voltage is used in
comparison to the other groups and therefore a slightly lower SNR is expected
and observed. SNR values were within the ranges reported in literature (2, 9,
30-32). Furthermore, subjective image quality was considered good or excellent
in 93.7 % of the scans.

Numerous studies have explored the possibilities of reducing both CM volume

and tube voltage (33-37). To the best of our knowledge, no other study evaluated
a rule of thumb to customize both CM and tube voltage and simultaneously
individualize the protocol based on TBW in abdominal CT imaging. In this
randomized controlled trial, CM injection was individualized based on TBW and
no significant differences in attenuation were found between weight categories
≤ 80 kg and > 80 kg (figure 2). Awai et al. showed that LBW might be the more
reliable parameter to base the injection protocol on compared to TBW and BSA
(8). However, LBW must be calculated using the Boer or the James formula, the
first being preferred for heavier patients (28), and this may prove to be too time
consuming for daily clinical practice affecting daily clinical routine. Therefore,
considering both time and effort, TBW might be more practical. Future research
can be directed toward the role of LBW in individualizing scan and CM injection
protocols, while taking cost-effectiveness into account.

Nowadays, CT scans are performed at lower tube voltages and most of the
scanners incorporate techniques such as ATCM and automated tube voltage
selection into their systems, thereby providing an easy method to individualize
radiation dose while optimizing image quality. At present, newer CT scanners

54

Vol_BM_productie.indd 54 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

are capable of automatically adapting various scan parameters to individual

patients, while CM is most often administered in a one size fits all approach.
This contradiction is easily explained by the fact that CM administration
is still a manual, and therefore a more time consuming, procedure. A
connection between scanner and CM injector might be the solution to further
individualisation of protocols.

Limitations

This study has several limitations. First of all, in this single-centre study,
a difference was found between the number of men and women in the
different groups. As this is a randomized controlled trial, it can be attributed 3
to coincidence, but effects on outcome parameters cannot be ruled out. The
distribution of fatty tissue is known to be different between genders: women
in general have more fatty tissue than men and as fat contains fewer blood
vessels, it doesn’t play an important role in the distribution of CM. However,
we corrected for the difference in proportions of men in the four groups in the
statistical analysis. Second, even though cardiac output is an important factor in
CM administration, it was not taken into account in this study. Timing in portal
venous CT is of lesser importance compared to arterial phase scans and all
patients were hemodynamically stable. It was therefore assumed that cardiac
output was within normal physiological ranges in all patients.

Conclusion

The proposed 10-to-10 rule is an easily reproducible method for achieving

homogeneous enhancement of the liver in portal venous abdominal CT,
irrespective of patient TBW or tube voltage.

55

Vol_BM_productie.indd 55 02/02/2022 15:22:55

 Chapter 3

References

1. Kondo H, Kanematsu M, Goshima S, Tomita Y, Kim MJ, Moriyama N, et al. Body

size indexes for optimizing iodine dose for aortic and hepatic enhancement at
multidetector CT: comparison of total body weight, lean body weight, and blood
volume. Radiology. 2010;254(1):163-9.
2. Heiken JP, Brink JA, McClennan BL, Sagel SS, Crowe TM, Gaines MV. Dynamic
incremental CT: effect of volume and concentration of contrast material and patient
weight on hepatic enhancement. Radiology. 1995;195(2):353-7.
3. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
4. Kondo H, Kanematsu M, Goshima S, Watanabe H, Kawada H, Moriyama N, et al.
Body size indices to determine iodine mass with contrast-enhanced multi-detector
computed tomography of the upper abdomen: does body surface area outperform
total body weight or lean body weight? Eur Radiol. 2013;23(7):1855-61.
5. Kondo H, Kanematsu M, Goshima S, Watanabe H, Onozuka M, Moriyama N, et al.
Aortic and hepatic enhancement at multidetector CT: evaluation of optimal iodine
dose determined by lean body weight. Eur J Radiol. 2011;80(3):e273-7.
6. Bae KT, Shah AJ, Shang SS, Wang JH, Chang S, Kanematsu M, et al. Aortic and hepatic
contrast enhancement with abdominal 64-MDCT in pediatric patients: effect of body
weight and iodine dose. AJR Am J Roentgenol. 2008;191(5):1589-94.
7. Ho LM, Nelson RC, Delong DM. Determining contrast medium dose and rate on
basis of lean body weight: does this strategy improve patient-to-patient uniformity
of hepatic enhancement during multi-detector row CT? Radiology. 2007;243(2):431-7.
8. Awai K, Kanematsu M, Kim T, Ichikawa T, Nakamura Y, Nakamoto A, et al. The optimal
body size index with which to determine iodine dose for hepatic dynamic CT: a
prospective multicenter study. Radiology. 2016;278(3):773-81.
9. Martens B, Hendriks BMF, Eijsvoogel NG, Wildberger JE, Mihl C. Individually body
weight-adapted contrast media application in computed tomography imaging of
the liver at 90 kVp. Invest Radiol. 2019;54(3):177-82.
10. Lell MM, Wildberger JE, Alkadhi H, Damilakis J, Kachelriess M. Evolution in computed
tomography: the battle for speed and dose. Invest Radiol. 2015;50(9):629-44.
11. Lell MM, Kachelriess M. Recent and upcoming technological developments in
computed tomography: High speed, low dose, deep learning, multienergy. Invest
Radiol. 2020;55(1):8-19.
12. Fleischmann U, Pietsch H, Korporaal JG, Flohr TG, Uder M, Jost G, et al. Impact of
contrast media concentration on low-kilovolt computed tomography angiography:
a systematic preclinical approach. Invest Radiol. 2018;53(5):264-70.
13. Dyer R, DiSantis DJ, McClennan BL. Simplified imaging approach for evaluation of
the solid renal mass in adults. Radiology. 2008;247(2):331-43.

56

Vol_BM_productie.indd 56 02/02/2022 15:22:55

 A Solution for Homogeneous Liver Enhancement in CT

14. Kang SK, Huang WC, Pandharipande PV, Chandarana H. Solid renal masses: what
the numbers tell us. AJR Am J Roentgenol. 2014;202(6):1196-206.
15. Kok M, Mihl C, Hendriks BM, Altintas S, Kietselaer BL, Wildberger JE, et al. Optimizing
contrast media application in coronary CT angiography at lower tube voltage:
Evaluation in a circulation phantom and sixty patients. Eur J Radiol. 2016;85(6):1068-
74.
16. Hendriks BMF, Eijsvoogel NG, Kok M, Martens B, Wildberger JE, Das M. Optimizing
pulmonary embolism computed tomography in the age of individualized medicine:
a prospective clinical study. Invest Radiol. 2018;53(5):306-12.
17. Eijsvoogel NG, Hendriks BMF, Willigers JL, Martens B, Carati LF, Horehledova B,
et al. Personalization of injection protocols to the individual patient’s blood
volume and automated tube voltage selection (ATVS) in coronary CTA. PLoS One.
2018;13(9):e0203682.
18. Canstein C, Korporaal JG. Reduction of contrast agent dose at low kV settings.
3
In: Siemens Healthineers White Paper, 2015: https://pdfs.semanticscholar.org/
ddf9/770a5467673d40025446b29707ef4bb26fbc.pdf.
19. Martens B, Hendriks BMF, Mihl C, Wildberger JE. Tailoring contrast media protocols
to varying tube voltages in vascular and parenchymal CT imaging: The 10-to-10 rule.
Invest Radiol. 2020;55:epub.
20. Germain T, Favelier S, Cercueil JP, Denys A, Krause D, Guiu B. Liver segmentation:
practical tips. Diagnostic and interventional imaging. 2014;95(11):1003-16.
21. Shuman WP, Chan KT, Busey JM, Mitsumori LM, Choi E, Koprowicz KM, et al.
Standard and reduced radiation dose liver CT images: adaptive statistical iterative
reconstruction versus model-based iterative reconstruction-comparison of findings
and image quality. Radiology. 2014;273(3):793-800.
22. Szucs-Farkas Z, Strautz T, Patak MA, Kurmann L, Vock P, Schindera ST. Is body weight
the most appropriate criterion to select patients eligible for low-dose pulmonary
CT angiography? Analysis of objective and subjective image quality at 80 kVp in 100
patients. Eur Radiol. 2009;19(8):1914-22.
23. Scholtz JE, Wichmann JL, Husers K, Beeres M, Nour-Eldin NE, Frellesen C, et al.
Automated tube voltage adaptation in combination with advanced modeled
iterative reconstruction in thoracoabdominal third-generation 192-slice dual-source
computed tomography: effects on image quality and radiation dose. Acad Radiol.
2015;22(9):1081-7.
24. Zhang X, Li S, Liu W, Huang N, Li J, Cheng L, et al. Double-low protocol for hepatic
dynamic CT scan: effect of low tube voltage and low-dose iodine contrast agent on
image quality. Medicine (Baltimore). 2016;95(26):e4004.
25. Song JS, Lee JM, Sohn JY, Yoon JH, Han JK, Choi BI. Hybrid iterative reconstruction
technique for liver CT scans for image noise reduction and image quality
improvement: evaluation of the optimal iterative reconstruction strengths. Radiol
Med. 2015;120(3):259-67.

57

Vol_BM_productie.indd 57 02/02/2022 15:22:55

 Chapter 3

26. Hendriks BM, Kok M, Mihl C, Bekkers SC, Wildberger JE, Das M. Individually
tailored contrast enhancement in CT pulmonary angiography. Br J Radiol.
2016;89(1061):20150850.
27. Tawfik AM, Kerl JM, Bauer RW, Nour-Eldin NE, Naguib NN, Vogl TJ, et al. Dual-energy
CT of head and neck cancer: average weighting of low- and high-voltage acquisitions
to improve lesion delineation and image quality-initial clinical experience. Invest
Radiol. 2012;47(5):306-11.
28. Caruso D, De Santis D, Rivosecchi F, Zerunian M, Panvini N, Montesano M, et al.
Lean body weight-tailored iodinated contrast injection in obese patient: boer versus
james formula. Biomed Res Int. 2018;2018:8521893.
29. Holmquist F, Soderberg M, Nyman U, Falt T, Siemund R, Geijer M. Can iterative
reconstruction algorithms replace tube loading compensation in low kVp
hepatic CT? Subjective versus objective image quality. Acta Radiol Open.
2020;9(3):2058460120910575.
30. Goshima S, Kanematsu M, Noda Y, Kawai N, Kawada H, Ono H, et al. Minimally
required iodine dose for the detection of hypervascular hepatocellular carcinoma
on 80-kVp CT. AJR Am J Roentgenol. 2016;206(3):518-25.
31. Goshima S, Kanematsu M, Noda Y, Kondo H, Watanabe H, Kawada H, et al.
Determination of optimal intravenous contrast agent iodine dose for the detection
of liver metastasis at 80-kVp CT. Eur Radiol. 2014;24(8):1853-9.
32. Nakamoto A, Kim T, Hori M, Onishi H, Tsuboyama T, Sakane M, et al. Clinical
evaluation of image quality and radiation dose reduction in upper abdominal
computed tomography using model-based iterative reconstruction; comparison
with filtered back projection and adaptive statistical iterative reconstruction. Eur J
Radiol. 2015;84(9):1715-23.
33. Brehmer K, Brismar TB, Morsbach F, Svensson A, Stal P, Tzortzakakis A, et al. Triple
arterial phase CT of the liver with radiation dose equivalent to that of single arterial
phase CT: Initial experience. Radiology. 2018;289(1):111-8.
34. Nakamoto A, Yamamoto K, Sakane M, Nakai G, Higashiyama A, Juri H, et al. Reduction
of the radiation dose and the amount of contrast material in hepatic dynamic
CT using low tube voltage and adaptive iterative dose reduction 3-dimensional.
Medicine (Baltimore). 2018;97(34):e11857.
35. Araki K, Yoshizako T, Yoshida R, Tada K, Kitagaki H. Low-voltage (80-kVp)
abdominopelvic computed tomography allows 60% contrast dose reduction in
patients at risk of contrast-induced nephropathy. Clin Imaging. 2018;51:352-5.
36. Buls N, Van Gompel G, Van Cauteren T, Nieboer K, Willekens I, Verfaillie G, et al.
Contrast agent and radiation dose reduction in abdominal CT by a combination
of low tube voltage and advanced image reconstruction algorithms. Eur Radiol.
2015;25(4):1023-31.

58

Vol_BM_productie.indd 58 02/02/2022 15:22:56

 A Solution for Homogeneous Liver Enhancement in CT

37. Nakaura T, Nakamura S, Maruyama N, Funama Y, Awai K, Harada K, et al. Low

contrast agent and radiation dose protocol for hepatic dynamic CT of thin adults at
256-detector row CT: effect of low tube voltage and hybrid iterative reconstruction
algorithm on image quality. Radiology. 2012;264(2):445-54.

59

Vol_BM_productie.indd 59 02/02/2022 15:22:56

Vol_BM_productie.indd 60 02/02/2022 15:22:57
 CHAPTER
4
Finding the optimal tube current
and iterative reconstruction strength
in liver imaging; two needles
in one haystack

B. Martens, J.G.A. Bosschee, S.M.J. Van Kuijk, C.R.L.P.N. Jeukens, M.T.H. Brauer,
J.E. Wildberger, C. Mihl

Submitted for publication

Vol_BM_productie.indd 61 02/02/2022 15:22:58

 Chapter 4

Abstract

Objectives: The aim of the study was to find the lowest possible tube current
and the optimal iterative reconstruction (IR) strength in abdominal imaging.

Material and Methods: Reconstruction software was used to insert noise,

simulating as if a lower tube current was used. An abdominal phantom was
used to validate the performance of the ReconCT software. Thirty abdominal
CT scans performed with a standard protocol (120 kVref, 150 mAsref ) scanned
at 90 kV, with dedicated contrast media (CM) injection software were selected.
The software was used to insert noise as if the scans were performed with
90, 80, 70 and 60 % of the full dose. Consequently, the different scans were
reconstructed with Filtered back projection (FBP) and IR strength 2, 3 and 4.
Objective and subjective image quality were evaluated. Based on the results,
lesion detection was graded by two radiologists in consensus in another 30
scans (identical scan protocol) with various liver lesions, reconstructed with IR
3, 4 and 5.

Results: A tube current of 60% still led to sufficient objective image quality when
IR strength 3 or 4 were used. IR strength 4 was preferred for lesion detection.
The subjective image quality was rated highest for the scans performed at
90% with IR 4.

Conclusion: A tube current reduction of 10 - 40% is possible in case IR 4 is used,

leading to the highest image quality (10%) or still sufficient image quality (40%),
shown by a pairwise comparison in the same patients.

Key words

Multidetector Computed Tomography; Diagnostic Imaging; Abdomen; Radiation

Dose Reduction; iterative reconstructions

62

Vol_BM_productie.indd 62 02/02/2022 15:22:58

 Optimal tube current and IR strength in liver imaging

Introduction

Computed tomography (CT) is used on a daily basis in abdominal imaging to

detect and evaluate a variety of pathologies (1, 2). While its clinical importance
and benefits are undisputed, CT uses ionizing radiation, which for radiation
exposure levels may lead to an increase of the lifetime attributable cancer
risk (3, 4). Radiation exposure should therefore be kept “as low as reasonably
achievable” according to the ALARA principle (5). Conversely, excessive radiation
dose reduction can lead to insufficient diagnostic image quality, resulting in
non-diagnostic scans. From a radiation safety point of view this is the worst-
case scenario as no medical benefit was gained from the radiation exposure
and a retake will lead in summary to an increase of the total radiation exposure.
This is indicative of the delicate balance between radiation dose and image
quality.
4

Several dose reduction techniques are applied in daily clinical practice, such as
automated tube current modulation (ATCM), automated tube voltage selection
(ATVS) and iterative reconstruction (IR) techniques (6-15). ATCM and ATVS
optimize radiation dose by optimizing the tube current and tube voltage during
the acquisition to reach sufficient image quality for each individual patient
(8, 10). IR techniques are used during reconstruction of the scans to further
decrease image noise, without compromising image quality (9). Based on the
same raw data IR techniques result in a decrease in image noise, by repetitive
calculation steps during the reconstruction. The repetition is stopped when a
predefined number of cycles is reached, or when the difference between two IR
steps becomes smaller than a predefined amount (16). The achievable decrease
in image noise, which is related to the IR strength, can be relinquished in favour
of a radiation dose reduction (17). Previous studies have shown IR techniques
to be superior to filtered back projection (FBP). Although, while Hardie et al.
showed a reader preference for low to intermediate IR strengths, Choy et al.
demonstrated a preference for images reconstructed with IR strength 4 or 5
(18, 19). Noise decreases with an increased IR strength, but at the same time,
the texture of the noise changes, possibly negatively influencing image quality
(20). Therefore, in daily clinical practice, different IR strengths are used (18,
21-23).

63

Vol_BM_productie.indd 63 02/02/2022 15:22:58

 Chapter 4

Comparing image quality between CT scans of different patients is challenging,

because differences in patient-related factors (e.g. height, weight, liver
morphology and cardiac function) affect image quality (6, 15). Ideally, if one
patient could be scanned several times, a reliable search for the most optimal
tube current and tube voltage could be performed. Reconstruction software
allows to reconstruct multiple lower tube current scans from a single raw
data set. Therefore, it provides the opportunity for pairwise comparison of
identical patients without the need for repetitive scanning. This software can
aid in finding the optimal reference tube current and help in further decreasing
radiation exposure. Previous studies showed that using dedicated post-
processing software (ReconCT) for optimization a potential dose reduction of
41 to 84 % was possible in CT angiography (CTA) of various vascular structures
in head and neck without compromising diagnostic image quality (24, 25).
The pairwise comparison stipulates the opportunity to evaluate in abdominal
imaging whether a dose reduction still leads to sufficient image quality and
lesion detection. The latter being one of the most important parameters, as
an increased risk of missing lesions is an unfavourable outcome. In a previous
study, signal to noise ratio (SNR) and contrast to noise ratio (CNR) values of
respectively 8.8 +/- 1.8 and 5.5 +/- 2.1 led to a good to excellent subjective image
quality in 93.7 % of the patients (26). Therefore, it is safely to assume that a SNR
above 8.0 and a CNR above 5.0 are considered sufficient.

The aim of this study was to assess both the optimal IR strength and the lowest
possible tube current in abdominal CT imaging while maintaining diagnostic
image quality with the use of ReconCT software.

Materials and Methods

Ethical considerations

This study was provided a waiver of written informed consent by the local
ethical committee and institutional review board as retrospective data were
analysed anonymously (ref METC 2017-0250).

64

Vol_BM_productie.indd 64 02/02/2022 15:22:58

 Optimal tube current and IR strength in liver imaging

Study design

ReconCT software (version 13.0.0.1, prototype software, Siemens Healthineers,

Forchheim, Germany) was used to reconstruct raw image data at lower tube
currents and different IR strengths to simulate radiation dose reduction and
accompanying image quality, i.e. an increase in noise. The raw CT image data
were exported directly from the CT-scanner. All scans were performed on
a 3th generation dual-source CT (DSCT) scanner (Somatom Force, Siemens
Healthineers, Forchheim, Germany). Validation of the software has been
published elsewhere (24, 25). Although, a quality assurance has been performed
in a phantom study (see Appendix).

Thirty abdominal scans were retrospectively selected and reconstructed

multiple times with lower mAs percentages and IR strengths. Both objective
4
and subjective image quality were evaluated. In addition, 30 abdominal scans all
containing a diversity of liver lesions, were retrospectively selected to determine
the optimal IR strength for lesion detection based on the results of the previous
steps.

Patient study

Thirty abdominal scans of unique patients in portal venous phase were included
between September 2019 and February 2020. Inclusion criteria were scans in
which ATCM (CareDose 4D; Siemens) and ATVS (CARE kV; Siemens) techniques
were used, with a reference tube voltage and tube current of respectively 120
kVref and 150 mAsref, with a slice collimation of 192 x 0.6 mm and gantry rotation
time 0.5 seconds. Only scans acquired at 90 kV in which a dosing factor of 0.4
g I/kg contrast media (CM) was used, were included to ensure a homogeneous
database. Dedicated CM injection software was used (P3T; Bayer Healthcare,
Berlin, Germany), which calculates CM volume and flow rate, based on the
linear relationship between body weight and injection duration (15). A history
of liver disease or surgery was not a reason for exclusion. General exclusion
criteria for a contrast-enhanced abdominal CT were applied (e.g. pregnancy,
renal insufficiency [estimated glomerular filtration rate < 30 mL/min per 1.73
m2] and iodine allergy).

65

Vol_BM_productie.indd 65 02/02/2022 15:22:58

 Chapter 4

The raw data of the selected scans were transferred to the ReconCT
workstation, where the scans reconstructed with lower tube currents were
simulated. As ATCM and ATVS techniques were used, the mAseff differed
between patients, this mAseff will hereinafter be referred to as the initial value.
Data were reconstructed with tube current of 90 %, 80 %, 70 % and 60 % of the
initial value. Based on the phantom study, reconstructions with a tube current
below 60 % were expected to be of insufficient image quality, and therefore,
these data were not simulated. In addition, all scans were reconstructed with
FBP and IR strengths 2, 3 and 4 (Advanced Modeled Iterative Reconstruction
[ADMIRE], Siemens Healthineers, Forchheim, Germany), kernel Br40. This
resulted in a total of 510 CT series (17 reconstructions for each patient).

Patients’ weight was asked prior to the CT scan and together with patient’ sex,
age and radiation dose information (e.g. mean effective mAs [mAseff ], CT dose
index [CTDIvol, in mGy] and dose length product [DLP, in mGy*m]) collected
from the PACS workstation (IMPAX version 6.6.1.5003, AGFA HealthCare N.V.,
Mortsel, Belgium). The CM volume (in ml), total iodine load (TIL, g I), flow rate
and iodine delivery rate (IDR, in g I/s) were monitored with a dedicated data
acquisition program (CertegaTM Informatics Solution; Bayer).

Image analysis

Data were transferred to the radiology workstation (SyngoViaTM, VB30; Siemens

Healthineers, Erlangen, Germany). In all reconstructions the mean Hounsfield
Units (HU) and standard deviation (SD) were measured by placing the largest
possible region of interest (ROI) in three different liver segments (area ≥ 1 cm2),
preferably segments 2, 5 and 8 (according to the Couinaud distribution), not
containing vessels, biliary ducts or regional anomalies (e.g. cysts, metastasis or
changes related to surgery) (27). The signal to noise ratio (SNR) was calculated
by dividing the mean HU of the liver by its SD. The difference between the mean
liver HU and the attenuation of the left paraspinal muscle, divided by the SD of
the paraspinal muscle resulted in the contrast to noise ratio (CNR).

66

Vol_BM_productie.indd 66 02/02/2022 15:22:58

 Optimal tube current and IR strength in liver imaging

Subjective image quality

Two radiologists (B.M. and C.M.) with respectively 4- and 9-years’ experience
in abdominal imaging rated all scans in consensus on diagnostic screens, while
being blinded to the simulated tube current and reconstruction method used.
The radiologists were allowed to adjust window levels. The overall image quality
and lesion detection capability were separately rated on a 5-point Likert scale
(1 very poor, 2 = poor, 3 = moderate, 4 = good, 5 = excellent) (15, 28). In search
for optimal image quality, scans rated as “good” or “excellent” were considered
of sufficient image quality. The simulated scans with the highest percentage of
scans with good or excellent image quality, were rated best.

Liver lesions
4
In addition to the previous patient study, 30 abdominal scans containing a
diversity of liver lesions (e.g. non-specific, benign or malignant) were collected
between June and August 2020. Scans were used for the evaluation of lesion
detection as the presence of actual lesions, makes it easier and more reliable
to evaluate this parameter. Scans were eligible for inclusion when the same
scan and CM injection protocol as in the patient study was used. IR strength
3, 4 and 5 were reconstructed on the scanner, based on the results of the first
patient study. Two radiologists (B.M. and C.M.) evaluated in consensus which
IR strength resulted in the best liver lesions detectability. The readers had to
choose the preferred strength out of the three reconstructed IR strengths.
The IR strength rated most often as best for lesion detection, was declared
the favoured strength.

Statistical analysis

Summaries of categorical variables were expressed as absolute numbers with

percentages and continuous variables as mean ± SD. A linear mixed-effects
model was used to account for the fact that of each patient 17 reconstructions
are made and hence, data were correlated. The effective mAs was added to
the model as covariate, with HU as dependent variable. The generalized mixed-
effects model with binomial link function was used to investigate if there was
an association between the IR strength and the dichotomized subjective

67

Vol_BM_productie.indd 67 02/02/2022 15:22:59

 Chapter 4

image quality (diagnostic image quality and lesion detection). Results of the
generalized linear mixed-effects model were expressed as odds ratio (OR) and
95% confidence interval (CI). Statistical software (SPSS, version 26.0; IBM Corp,
New York, NY) was used for the data analysis.

Results

Table 1. Baseline characteristics.

Parameters (N N = 29
Age (years) 64.9 ± 14.2
Sex (% male) 18 (62.1 %)
Body weight (kg) 72.2 ± 9.9
Height (m) 1.73 ± 0.1
BMI (kg/m ) -2
24.2 ± 2.4
BMI indicates body mass.

Patient study

Baseline characteristics of the population are depicted in table 1. Mean mAseff,

CTDIvol, DLP, CM volume, TIL, flow rate and IDR are shown in table 2. Data from
one patient was excluded, as a higher dosing factor (in g I/kg) was used.

Table 2. Radiation dose and injection parameters.

Parameters N = 29
Mean mAseff 212.0 ± 27.9
CTDIvol (mGy) 6.1 ± 0.8
DLP (mGy*cm) 291.4 ± 43.8
CM volume (ml) 95.8 ± 13.1
TIL (g) 28.8 ± 3.9
Flow rate (ml/s) 2.9 ± 0.6
IDR (g I/s) 0.96 ± 0.1
Dosing factor (g I/kg) 0.4
mAseff indicates effective tube current; CTDIvol, CT dose index vol; DLP, dose length product;
CM, contrast media; TIL, total iodine load; IDR, iodine delivery rate.

68

Vol_BM_productie.indd 68 02/02/2022 15:22:59

 Optimal tube current and IR strength in liver imaging

To find the optimal tube current and IR strength, SNR and CNR were evaluated
(figure 1). Figure 1a and 1b show in green which percentage in mAs reduction
still leads to a sufficient SNR and CNR. In case IR strength 3 or 4 is used, a mAs
of 60 % still results in sufficient SNR and CNR. Figure 1c depicts the overall
image quality with each reconstruction strength. The odds that IR strength 3
results in a diagnostic scan was eight times higher than that of FBP

Figure 1. A signal to noise ratio (SNR) of 8.0 (A) and contrast to noise ratio (CNR) of
5.0 (B) were considered sufficient. A and B show the corresponding SNR and CNR for
each combination of iterative reconstruction (IR) strength and percentage of the initial
value of the tube current. In green the combination leading to sufficient objective image
quality. In part C and D, the odds ratios of the overall diagnostic image quality (C) and
the lesion detection capability (D) are set out. Filtered back projection (FBP) and IR
strengths on the left are compared to the reconstruction methods on the x-axis. For
example, the odds that IR 4 results in a better lesion detection than IR 3 is 1.2, with a
confidence interval (CI) of 0.4-3.7. 4

69

Vol_BM_productie.indd 69 02/02/2022 15:22:59

 Chapter 4

and more than two times higher than strength 2 and 4. The odds that IR 4
results in a better lesion detection was 7.5 times higher than that of FBP and
respectively 1.3 and 1.2 times higher than that of IR 2 and IR 3 (figure 1d).

Figure 2. Abdominal scan of an 86-year-old patient (Patient A) in the follow-up for a

urothelial cell carcinoma, who has multiple cysts in the liver parenchyma. In addition, a
scan of a 47-year-old male (Patient B) in the follow-up for hepatic metastasis of colorec-
tal cancer. Both scans are reconstructed with IR strength 3, 4 and 5. The scans recon-
structed with IR 4 were rated in consensus to have the best lesion detection capability

70

Vol_BM_productie.indd 70 02/02/2022 15:22:59

 Optimal tube current and IR strength in liver imaging

A mAs of 60 % with the use of IR strength 3 or 4, still leads to sufficient objective

image quality. The overall subjective diagnostic image quality was highest for
IR 3. IR 4 was graded best for lesion detection.

The percentage of scans considered of sufficient diagnostic image quality (rated

as of good or excellent diagnostic image quality) was highest (89.7 %) for the
scan at 100 % with IR 2. With IR 3 at 90, 80, 70 and 60 % respectively 79.3,
69.0, 65.5 and 48.3 % of the scans was rated sufficient. Sufficient diagnostic
image quality was reached in 82.8, 72.4, 79.3 and 55.2 %, respectively at IR
4. Regarding lesion detection, the percentage of sufficient scans was 79.3 %
at 100 % with IR 2, while at IR 3 the percentages at 90, 80, 70 and 60 % were
respectively 72.4, 62.1, 55.2 and 37.9 %. At IR 4 the percentage of scans rated
as excellent or good was 86.2, 69.0, 69.0 and 55.2 % respectively.
4
Liver lesions

This second patient population confirmed the preference for IR strength

4 regarding lesions detection. In twenty-five cases IR strength 4 was most
appreciated, while IR 3 was valued highest in 4 cases and IR 5 in only one case.
Examples of two cases in which IR 4 was preferred are depicted in figure 2.

71

Vol_BM_productie.indd 71 02/02/2022 15:22:59

 Chapter 4

Discussion

The aim of the study was to find the optimal IR strength and the lowest possible
(reference) tube current that could be used in abdominal CT imaging, without
compromising objective and subjective image quality. In accordance with the
literature, IR techniques outperformed FBP (18, 19, 22). When IR techniques
are used, the mAsref can be reduced without compromising objective image
quality. The results indicate that with IR strength 3 or 4, reductions of up to 40
% still produce a sufficient SNR and CNR. Scans performed with IR 4 at 90 %
tube current, led to a slightly higher lesion detection capability compared to
the full dose at IR strength 2. Therefore, it can be concluded that the mAsref in
abdominal imaging can be safely reduced by 10 – 40 %, in case IR strength 4 is
used on this particular scanner, showed by pairwise comparison. Ten percent
reduction at IR 4 leads to the highest image quality, while a reduction of 40 %
at IR strength 4 still results in sufficient image quality.

For the first patient study, only scans with IR strength 2 to 4 were reconstructed.
IR strength 1 and 5 were not reconstructed. From experience, IR 1 was expected
to result in very noisy images and IR 5 in images appearing very smoothened.
IR 4 turned out to result in subjectively the best lesion detection capability.
Subsequently, the second study was performed, in which IR 5 was incorporated
in addition to IR strength 3 and 4 to rule out possible superiority of IR strength
5.

A number of studies have evaluated the possibility to reduce radiation dose

in abdominal imaging (23, 29-33). To the best of our knowledge, this is the
first study comparing different radiation doses and IR strengths in abdominal
imaging, within the same patient by using reconstruction software. The study
set up can be used to investigate the optimal tube current and IR algorithm for
each anatomical region, scan indication, vendor and specific scanner.

Our study evaluated both objective and subjective image quality. For the latter,
as the name already implies, it is subjective and some readers might prefer
more noise for a particular scan indication, while others prefer smoothened
scans (17). Establishing the objective image quality with SNR and CNR seems
rather straightforward. Although, when searching for reliable thresholds, a wide

72

Vol_BM_productie.indd 72 02/02/2022 15:22:59

 Optimal tube current and IR strength in liver imaging

variety of values is found in the literature, all presumed to be of diagnostic

image quality, and no clear cut off values are established (33-37). In addition,
previous literature states that both SNR and CNR might not encompass the
complete appreciation of image quality (38, 39). The present study sets out
the discrepancy between objective and subjective image quality. According to
the SNR and CNR values a radiation dose reduction of 40 % was possible, while
looking at the subjective parameters only a smaller tube current reduction
of 10 - 40 % seemed feasible. This indicates the struggle to be able to safely
declare that CT scans are of sufficient image quality. Future research could
focus on determining new, more universal objective parameters to reliable,
generalizable and consequently assess image quality. Such parameters would
make it possible to establish with a higher degree of certainty if image quality is
diagnostic and if all the different developed radiation dose reduction algorithms
and reconstructions results in sufficient image quality for diagnostic purposes.
4

Limitations

The study is a single-center study with a rather small patient sample. In addition,
the golden standard for lesion detection and characterization is autopsy, which
was not performed. The baseline protocol for abdominal imaging chosen in
present study was the scan and CM injection protocol as used in daily clinical
practice. This assumes that this baseline scan protocol is considered to be
of good – maybe even too good - image quality, while this protocol might
potentially have benefitted from a (small) increase in dose. Lastly, radiation
dose reduction and IR strengths were only studied on a CT scanner from one
vendor, which limits generalizability of the outcome. As the software is vendor
specific and raw data based it is therefore not applicable to scanners form
other vendors.

Conclusion

IR strength 4 leads to the best subjective image quality in abdominal CT imaging

and gives the opportunity to reduce the tube current by 10 to 40 % without
compromising both objective as well as subjective image quality as shown by
pairwise comparison in the same patients with the use of ReconCT software.

73

Vol_BM_productie.indd 73 02/02/2022 15:22:59

 Chapter 4

References

1. Coskun M. Hepatocellular carcinoma in the cirrhotic liver: Evaluation using computed

tomography and magnetic resonance imaging. Exp Clin Transplant. 2017;15(Suppl
2):36-44.
2. Chou R, Cuevas C, Fu R, et al. Imaging techniques for the diagnosis of hepatocellular
carcinoma: A systematic review and meta-analysis. Ann Intern Med. 2015;162(10):697-
711.
3. The 2007 recommendations of the international commission on radiological
protection. Icrp publication 103. Ann ICRP. 2007;37(2-4):1-332.
4. Board of Radiation Effects Research Division on Earth and Life Sciences National
Research Council of the National Academies. Health risks from exposure to low
levels of ionizing radiation: Beir vii, phase 2. Washington (DC); 2006.
5. Barrett B, Stiles M, Patterson J. Radiation risks: Critical analysis and commentary.
Prev Med. 2012;54(3-4):280-2.
6. Kondo H, Kanematsu M, Goshima S, et al. Body size indexes for optimizing iodine
dose for aortic and hepatic enhancement at multidetector ct: Comparison of total
body weight, lean body weight, and blood volume. Radiology. 2010;254(1):163-9.
7. Lell MM, Wildberger JE, Alkadhi H, et al. Evolution in computed tomography: The
battle for speed and dose. Invest Radiol. 2015;50(9):629-44.
8. Kalra MK, Maher MM, Toth TL, et al. Techniques and applications of automatic tube
current modulation for ct. Radiology. 2004;233(3):649-57.
9. Arapakis I, Efstathopoulos E, Tsitsia V, et al. Using “idose4” iterative reconstruction
algorithm in adults’ chest-abdomen-pelvis ct examinations: Effect on image quality
in relation to patient radiation exposure. Br J Radiol. 2014;87(1036):20130613.
10. Martin CJ, Sookpeng S. Setting up computed tomography automatic tube current
modulation systems. J Radiol Prot. 2016;36(3):R74-r95.
11. Lell MM, Kachelriess M. Recent and upcoming technological developments in
computed tomography: High speed, low dose, deep learning, multienergy. Invest
Radiol. 2020;55(1):8-19.
12. Papadakis AE, Damilakis J. Automatic tube current modulation and tube voltage
selection in pediatric computed tomography: A phantom study on radiation dose
and image quality. Invest Radiol. 2019;54(5):265-72.
13. Eller A, May MS, Scharf M, et al. Attenuation-based automatic kilovolt selection in
abdominal computed tomography: Effects on radiation exposure and image quality.
Invest Radiol. 2012;47(10):559-65.
14. De Cecco CN, Darnell A, Macias N, et al. Virtual unenhanced images of the abdomen
with second-generation dual-source dual-energy computed tomography: Image
quality and liver lesion detection. Invest Radiol. 2013;48(1):1-9.

74

Vol_BM_productie.indd 74 02/02/2022 15:22:59

 Optimal tube current and IR strength in liver imaging

15. Martens B, Hendriks BMF, Eijsvoogel NG, et al. Individually body weight-adapted
contrast media application in computed tomography imaging of the liver at 90 kvp.
Invest Radiol. 2019;54(3):177-82.
16. Stiller W. Basics of iterative reconstruction methods in computed tomography: A
vendor-independent overview. Eur J Radiol. 2018;109:147-54.
17. Geyer LL, Schoepf UJ, Meinel FG, et al. State of the art: Iterative ct reconstruction
techniques. Radiology. 2015;276(2):339-57.
18. Hardie AD, Nelson RM, Egbert R, et al. What is the preferred strength setting of
the sinogram-affirmed iterative reconstruction algorithm in abdominal ct imaging?
Radiol Phys Technol. 2015;8(1):60-3.
19. Choy S, Parhar D, Lian K, et al. Comparison of image noise and image quality between
full-dose abdominal computed tomography scans reconstructed with weighted
filtered back projection and half-dose scans reconstructed with improved sinogram-
affirmed iterative reconstruction (safire*). Abdom Radiol (NY). 2019;44(1):355-61.
20. Morsbach F, Desbiolles L, Raupach R, et al. Noise texture deviation: A measure for
quantifying artifacts in computed tomography images with iterative reconstructions.
4
Invest Radiol. 2017;52(2):87-94.
21. Choy S, Parhar D, Lian K, et al. Comparison of image noise and image quality between
full-dose abdominal computed tomography scans reconstructed with weighted
filtered back projection and half-dose scans reconstructed with improved sinogram-
affirmed iterative reconstruction (safire*). Abdominal Radiology. 2018;44.
22. Wang R, Schoepf UJ, Wu R, et al. Diagnostic accuracy of coronary ct angiography:
Comparison of filtered back projection and iterative reconstruction with different
strengths. J Comput Assist Tomogr. 2014;38(2):179-84.
23. Kataria B, Nilsson Althen J, Smedby O, et al. Assessment of image quality in
abdominal computed tomography: Effect of model-based iterative reconstruction,
multi-planar reconstruction and slice thickness on potential dose reduction. Eur J
Radiol. 2020;122:108703.
24. Ellmann S, Kammerer F, Brand M, et al. A novel pairwise comparison-based method
to determine radiation dose reduction potentials of iterative reconstruction
algorithms, exemplified through circle of willis computed tomography angiography.
Invest Radiol. 2016;51(5):331-9.
25. Kramer M, Ellmann S, Allmendinger T, et al. Computed tomography angiography of
carotid arteries and vertebrobasilar system: A simulation study for radiation dose
reduction. Medicine (Baltimore). 2015;94(26):e1058.
26. Martens B, Wildberger JE, Hendriks BMF, et al. A solution for homogeneous liver
enhancement in computed tomography: Results from the complex trial. Invest
Radiol. 2020;55(10):666-72.
27. Sibulesky L. Normal liver anatomy. Clin Liver Dis (Hoboken). 2013;2(Suppl 1):S1-s3.
28. Jamieson S. Likert scales: How to (ab)use them. Med Educ. 2004;38(12):1217-8.

75

Vol_BM_productie.indd 75 02/02/2022 15:22:59

 Chapter 4

29. Brehmer K, Brismar TB, Morsbach F, et al. Triple arterial phase ct of the liver with
radiation dose equivalent to that of single arterial phase ct: Initial experience.
Radiology. 2018;289(1):111-8.
30. Nakamoto A, Yamamoto K, Sakane M, et al. Reduction of the radiation dose and
the amount of contrast material in hepatic dynamic ct using low tube voltage
and adaptive iterative dose reduction 3-dimensional. Medicine (Baltimore).
2018;97(34):e11857.
31. Araki K, Yoshizako T, Yoshida R, et al. Low-voltage (80-kvp) abdominopelvic computed
tomography allows 60% contrast dose reduction in patients at risk of contrast-
induced nephropathy. Clin Imaging. 2018;51:352-5.
32. Buls N, Van Gompel G, Van Cauteren T, et al. Contrast agent and radiation dose
reduction in abdominal ct by a combination of low tube voltage and advanced image
reconstruction algorithms. Eur Radiol. 2015;25(4):1023-31.
33. Holmquist F, Soderberg M, Nyman U, et al. 80-kvp hepatic ct to reduce contrast
medium dose in azotemic patients: A feasibility study. Acta Radiol. 2020;61(4):441-9.
34. Goshima S, Kanematsu M, Noda Y, et al. Determination of optimal intravenous
contrast agent iodine dose for the detection of liver metastasis at 80-kvp ct. Eur
Radiol. 2014;24(8):1853-9.
35. Miyoshi K, Onoda H, Tanabe M, et al. Image quality in dual-source multiphasic
dynamic computed tomography of the abdomen: Evaluating the effects of a low
tube voltage (70 kvp) in combination with contrast dose reduction. Abdom Radiol
(NY). 2020;45(11):3755-62.
36. Akagi M, Nakamura Y, Higaki T, et al. Deep learning reconstruction improves image
quality of abdominal ultra-high-resolution ct. Eur Radiol. 2019;29(11):6163-71.
37. Choi SJ, Ahn SJ, Park SH, et al. Dual-source abdominopelvic computed tomography:
Comparison of image quality and radiation dose of 80 kvp and 80/150 kvp with tin
filter. PLoS One. 2020;15(9):e0231431.
38. Vaishnav JY, Jung WC, Popescu LM, et al. Objective assessment of image quality and
dose reduction in ct iterative reconstruction. Med Phys. 2014;41(7):071904.
39. De Crop A, Smeets P, Van Hoof T, et al. Correlation of clinical and physical-technical
image quality in chest ct: A human cadaver study applied on iterative reconstruction.
BMC Med Imaging. 2015;15:32.

76

Vol_BM_productie.indd 76 02/02/2022 15:22:59

 Optimal tube current and IR strength in liver imaging

77

Vol_BM_productie.indd 77 02/02/2022 15:22:59

Vol_BM_productie.indd 78 02/02/2022 15:23:01
 CHAPTER
5
Individualized scan protocols in
abdominal computed tomography:
radiation versus contrast media dose
optimization

B. Martens, G. Jost, C. Mihl, J.E. Wildberger, B. Schmidt, T. Flohr, H. Pietsch

Published in:
Investigative Radiology 2021; Epub ahead of print.

Vol_BM_productie.indd 79 02/02/2022 15:23:02

 Chapter 5

Abstract

Background: In contrast-enhanced abdominal computed tomography (CT),

radiation and contrast media (CM) injection protocols are closely linked to each
other and therefore a combination is the basis for achieving optimal image
quality. However, most studies focus on optimizing one or the other parameter
separately.

Purpose: Reducing radiation dose may be most important for a young patient
or a population in need of repetitive scanning, whereas CM reduction might
be key in a population with insufficient renal function. The recently introduced
technical solution, in the form of an automated tube voltage selection slider
[ATVS], might be helpful in this respect. The aim of the current study was to
systematically evaluate feasibility of optimizing either radiation or CM dose in
abdominal imaging compared to a combined approach.

Methods: Six Göttingen minipigs (mean weight 38.9 ± 4.8 kg) were scanned on
a 3rd-generation dual source CT. ATVS and automated tube current modulation
(ATCM) techniques were used, with quality reference values of 120 kVref and 210
mAsref. ATVS was set at 90 kV semimode. Three different abdominal scan and
CM protocols were compared intra-individually: 1. the standard “combined”
protocol, with the ATVS slider position set at 7 and a body weight adapted CM
injection protocol of 350 mg I/kg body weight, iodine delivery rate (IDR) of 1.1 g
I/s; 2. the CM dose saving protocol, with the ATVS slider set at 3 and CM dose
lowered to 294 mg I/kg, resulting in a lower IDR of 0.9 g I/s; 3. the radiation
dose saving protocol, with the ATVS slider position set at 11 and a CM dose of
441 mg I/kg and an IDR 1.3 g I/s, respectively. Scans were performed with each
protocol in arterial, portal venous and delayed phase. Objective image quality
was evaluated by measuring the attenuation in Hounsfield Units (HU), signal-
to-noise ratio (SNR), and contrast-to-noise ratio (CNR) of the liver parenchyma.
The overall image quality, contrast quality, noise and lesion detection capability
were rated on a 5-point Likert scale (1 = excellent, 5 = very poor). Protocols were
compared for objective image quality parameters using one-way ANOVA, and
for subjective image quality parameters using Friedman test.

80

Vol_BM_productie.indd 80 02/02/2022 15:23:02

 Individualized Abdominal CT Protocols

Results: Mean radiation doses were 5.2 ± 1.7 mGy for the standard protocol, 7.1
± 2.0 mGy for the CM dose saving protocol, and 3.8 ± 0.4 mGy for the radiation
dose saving protocol. Mean total iodine load (TIL) in these groups was 13.7 ± 1.7,
11.4 ± 1.4 and 17.2 ± 2.1 g, respectively. No significant differences in subjective
overall image or contrast quality were found. SNR and CNR were not significantly
different between protocols in any scan phase. Significantly more noise was
seen when using the radiation dose saving protocol (P < 0.01). In portal venous
and delayed phases, mean attenuation of the liver parenchyma significantly
differed between protocols (P < 0.001). Lesion detection was significantly better
in portal venous phase using the CM dose saving protocol compared to the
radiation dose saving protocol (P = 0.037).

Conclusion: In this experimental set-up, optimizing either radiation (-26 %) or

CM dose (-16 %) is feasible in abdominal CT imaging. Individualizing either
radiation or CM dose leads to comparable objective and subjective image
quality. Personalized abdominal CT examination protocols can thus be tailored
5
to individual risk assessment and might offer additional degrees of freedom.

Key words

Computed tomography; abdomen; liver; contrast media; radiation dose; renal

insufficiency; age; image quality; automated tube voltage selection

81

Vol_BM_productie.indd 81 02/02/2022 15:23:02

 Chapter 5

Introduction

Computed tomography (CT) of the abdomen is the workhorse of daily clinical

practice and is used for the diagnosis of a wide variety of pathologies (1). In
recent years, contrast media (CM) injection protocols have been individualized
based on different body size parameters (e.g. total body weight, body surface
area, and lean body weight) (2-10). Similarly, modern CT scanners automatically
individualize both tube current (automated tube current modulation; ATCM)
and tube voltage (automated tube voltage selection; ATVS) based on patient
body habitus. ATVS techniques are intended for contrast-enhanced CT scans,
because they exploit the strong increase of iodine attenuation at lower tube
voltage. Depending on patient body shape and imaging task, ATVS proposes
the tube voltage that provides a desired contrast-to-noise ratio (CNR) at
lowest radiation dose (11-13) – typically the lowest tube voltage with sufficient
tube current reserves for the planned examination. The extent of radiation
dose reduction at lower kV can be controlled by the user, e. g. by applying
different slider settings. In its vendor-recommended parametrization, ATVS
focuses on radiation dose reduction and assumes the same CM protocol is
used at all tube voltages. CNR, however, is a combination of both radiation
dose and iodine contrast. Therefore, by decreasing radiation dose beyond the
proposed ATVS parameters (e. g. by deviating from the vendor-recommended
slider settings), and at the same time increasing CM dose (or vice versa), similar
CNR’s can be reached (5). This offers perspective for further individualization of
radiation and CM protocols. For example, in younger patients and/or in patients
requiring repetitive scanning, a protocol favoring radiation dose reduction is
preferred over a decrease in CM dose, so as to minimize the increase in lifetime
attributable cancer risk due to ionizing radiation exposure (14-16). On the other
hand, in the elderly where reduced renal function is more common, a decrease
in CM dose is preferred over radiation dose reduction (17). Both radiation
dose and CM injection protocols can be manually adapted, but the slider bar
provided in ATVS to tailor the scan protocol offers a user-friendly alternative.

The current study aims to evaluate the feasibility of using standard ATVS slider
positions combined with adapted CM injection protocols for reducing either
radiation or CM dose, depending on individual risk assessment, compared
to a standard combined protocol. This was done by structurally comparing

82

Vol_BM_productie.indd 82 02/02/2022 15:23:02

 Individualized Abdominal CT Protocols

objective and subjective image quality parameters in imaging of the abdomen

within and between subjects.

Materials and Methods

Animals

The study was performed on 6 healthy female Göttingen minipigs (Ellegaard,

Dalmose, Denmark) with a mean body weight of 38.9 ± 4.8 kg. Three imaging
protocols were compared intra-individually in all 6 animals with at least one
week between examinations.

The animals were handled in compliance with the German Animal Welfare
Legislation and approval of the State Animal Welfare Committee. All
measurements were performed under general anesthesia and animals were
orally intubated and mechanically ventilated. Animals were placed in a prone
5
position and CT imaging was performed during end-expiratory ventilation stop.

Study design

CT imaging was performed on a third-generation dual source CT scanner

(Somatom Force, Siemens Healthineers, Forchheim, Germany). Abdominal
scans were done using slice collimation 192 × 0.6 mm, rotation time 0.5 s and
pitch 0.85, resulting in a scan time of 4 s. Image reconstruction was done with a
Br40 kernel, SAFIRE iterative reconstruction (level 3) at 0.75 mm slice thickness
with 0.5 mm increment. The ATVS system (CAREkV, Siemens Healthineers) was
operated with 90 kV semimode, ATCM and fixed quality reference values (120
kV, 210 mAs). The ATVS slider position is determined by the scan indication. For
parenchymal (e. g. liver) studies, the vendor recommends position 7 to balance
image noise increase and increased CM attenuation at lower tube voltage. The
standard protocol in this study was performed with this configuration (slider
position 7). At lower slider settings, less image noise increase is accepted at
lower tube voltage, with the consequence of higher radiation dose. Level 3,
originally intended for non-contrast scans, can therefore be used to perform
CT scans with similar CNR compared to level 7 but reduced CM volume. This
is the CM dose saving protocol used in present study. At slider position 11 -

83

Vol_BM_productie.indd 83 02/02/2022 15:23:02

 Chapter 5

originally intended for CT angiographic examinations - more image noise is

accepted at further reduced radiation dose. To maintain the expected CNR of
slider level 7, the CM volume needs to be increased. This is the radiation dose
saving protocol. The protocol specific CT scan configurations were combined
with adapted CM injection protocols. Iopromide (Ultravist 300, Bayer AG, Berlin,
Germany) was used and CM administration was performed with the Medrad
Centargo CT injection system (Bayer AG) into the ear vein of the animals. For
the standard imaging protocol 350 mg I/kg body weight was administered with
a flow rate of 3.5 ml/s, the Iodine delivery rate (IDR) was 1.1 g I/s. For the CM
dose saving protocol, the used standard dose (350 mgI/kg) was reduced by 16
% to 294 mg I/kg and flow rate was adapted to 2.9 ml/s (IDR 0.9 g I/s), so as to
maintain the same total injection time. A 26 % higher CM dose (441 mgI/kg)
than the used standard dose administered at 4.4 ml/s (IDR 1.3 g I/s) was used
for the low radiation dose protocol (18). All CM injections were followed by a 20
ml saline chaser applied at the same flow rate. A summary of the combination
of the scanner configuration and CM injections for each imaging protocol is
given in table 1.

Contrast timing was adjusted with bolus tracking in the descending aorta
using a threshold of 100 HU. Arterial phase imaging started with a delay of 5
s followed by the portal-venous and late phase using fixed delays of 60 s and
90 s.

The CTDI radiation doses were obtained from the dose reports of the CT
scanner. The percentage change in relation to the standard imaging protocol
was calculated.

Objective image quality

The data was evaluated on post-processing software (SyngoViaTM, VB30; Siemens

Healthineers, Erlangen, Germany). The Hounsfield Unit (HU) and standard
deviation (SD) were measured in the hepatic artery in arterial phase, by placing
a region of interest (ROI) as large as possible in the vascular structures, taking
into account the vasculature wall. In portal venous phase, three ROI’s (area ≥ 2
cm2) were drawn in three different liver segments. Preferably in segment 2, 5
and 8, according to the Couinaud distribution, not containing vessels or biliary

84

Vol_BM_productie.indd 84 02/02/2022 15:23:03

 Individualized Abdominal CT Protocols

ducts (19). Another as large as possible ROI was placed in the portal vein to
measure the signal attenuation. The SD of the paraspinal muscle (ROI area ≥
1 cm2) was used to estimate image noise. The signal-to-noise ratio (SNR) was
calculated by dividing the mean HU of the three liver segments by the noise.
The attenuation of the left paraspinal muscle was used to calculate the CNR.
Mean liver HU minus the HU of the paraspinal muscle, divided by the SD of the
paraspinal muscle resulted in the CNR. Similar calculations were performed for
the delayed phase.

Subjective image quality

The scans were rated in consensus on diagnostic screens by two radiologists

(C.M. and B.M.) with 10- and 5-years’ experience in abdominal imaging.
Adjusting the window level was allowed. The overall image quality, noise and
contrast quality were rated on a 5-point Likert scale (1 = excellent, 2 = good,
3 = moderate, 4 = poor, 5 = very poor) (7, 20). Lesion detection was rated in
5
portal venous and delayed phase using the same Likert scale. The arterial
phase is not solely used for liver lesion detection at our center, therefore this
parameter was not considered relevant.

Statistics

All results are presented as mean ± SD, or median with interquartile range
(IQR) for subjective image quality. Heart rate, attenuation, SNR and CNR
were compared between the three imaging protocols using one-way analysis
of variance on ranks (ANOVA) followed by the post-hoc Tukey’s multiple
comparisons test. Subjective image quality parameters were compared between
protocols using the Friedmann test followed by the Dunn’s test for multiple
comparison. Two-sided P values < 0.05 were regarded as statistically significant.
Statistical analyses were performed using GraphPad Prism (GraphPad Software
version 8, La Jolla, CA, USA).

85

Vol_BM_productie.indd 85 02/02/2022 15:23:03

 Chapter 5

Results

Injection parameters and radiation dose

Mean heart rates did not significantly differ between protocols: 104 ± 20 bpm
(standard), 105 ± 11 bpm (CM saving) and 102 ± 24 bpm (radiation saving). Table
1 shows an overview of radiation dose and CM injection parameters. As a result
of the study design, CM volumes and radiation doses differed between groups.
In the standard protocol, the CTDIvol, mean CM

Table 1. Contrast media (CM) and radiation dose parameters

Protocol
Standard CM dose saving Radiation dose
(n = 6) (n = 6) saving (n = 6)
Radiation dose parameters
CAREkV 90 kV semimode 90 kV semimode 90 kV semimode
Reference (kV/mAs) 120 / 210 120 / 210 120 / 210
Slider position 7 3 11
CTDIvol (mGy) 5.2 ± 1.7 7.1 ± 2.0 3.8 ± 0.4
CM injection parameters
Concentration (mg I/ml) iopromide 300 iopromide 300 iopromide 300
CM dose (mg I/kg) 350 294 441
Mean CM volume (ml) 45.5 ± 5.5 38 ± 4.8 57.3 ± 6.9
TIL (g) 13.7 ± 1.7 11.4 ± 1.4 17.2 ± 2.1
Flow rate (ml/s) 3.5 2.9 4.4
IDR (g I/s) 1.1 0.9 1.3
Saline chaser (ml) 20 20 20
TIL, total iodine load; IDR, iodine delivery rate (in g I/s); CTDI vol, CT dose index vol;

volume, and TIL were 5.2 ± 1.7 mGy, 45.5 ± 5.5 ml and 13.7 ± 1.7 g, respectively.
The mean radiation dose was higher in the CM dose saving group and lower
in the radiation dose saving group, with values of 7.1 ± 2.0 and 3.8 ± 0.4 mGy
respectively. The TIL was lowest in the CM dose saving group (11.4 ± 1.4 g) and
highest for the radiation dose saving group (17.2 ± 2.1 g).

86

Vol_BM_productie.indd 86 02/02/2022 15:23:03

 Individualized Abdominal CT Protocols

Objective image quality

Significant differences in attenuation (HU) of the hepatic artery in arterial phase,

and attenuation of the portal vein and liver in portal venous and the liver in
delayed phase were found, with P values < 0.001 in all cases (figure 1). Mean
attenuation of the liver parenchyma in portal venous phase was 130.6 ± 10.5
HU for the standard protocol. Attenuation was lower using the CM dose saving
protocol (121.3 ± 4.9 HU) and higher using the radiation dose saving protocol
(148.3 ± 6.3 HU) (P < 0.001).

Figure 1. Effect of contrast media (CM) and radiation dose protocols on mean atten-
uation of the hepatic artery, portal vein and liver parenchyma in three different scan
phases. Error bars indicate the standard deviation.

Abbreviations: HU: Hounsfield units

SNR and CNR did not significantly differ between groups in the arterial, portal
venous, or delayed phases (figure 2). Mean SNR of the liver in portal venous
phase was 8.2 ± 1.1 for the standard protocol, 9.8 ± 1.7 for the CM dose saving
protocol, and 8.6 ± 0.5 for radiation dose saving protocol (P= 0.188). Mean
CNR for the three protocols was 4.5 ± 1.3, 4.5 ± 1.0 and 4.5 ± 0.4, respectively
(P = 0.990) (table 2).

87

Vol_BM_productie.indd 87 02/02/2022 15:23:03

 Chapter 5

Figure 2. Effect of contrast media (CM) and radiation dose protocols on signal-to-noise
ratio (SNR) and contrast-to-noise ratio (CNR) of the hepatic artery, portal vein and liver
parenchyma in three different scan phases. Error bars indicate the standard deviation.

Table 2. Objective and Subjective image quality parameters

Protocol
Standard CM dose Radiation P
saving dose saving
Arterial phase
Objective image quality
Mean HU hepatic artery 627.5 ± 99.6 518.3 ± 75.2 788.5 ± 59.4 < 0.0011
Mean HU liver parenchyma 81.4 ± 11.2 81.6 ± 15.0 97.5 ± 25.3 0.267
SNR Liver 5.6 ± 1.0 6.0 ± 1.5 5.6 ± 1.8 0.838
CNR Liver 1.1 ± 1.0 1.3 ± 1.1 1.8 ± 1.5 0.66
Subjective image quality (median, IQR)
Overall 3 (2-3) 2.5 (2-3) 3 (3-3.3) 0.259
Noise 3 (2.8-3.3) 2.5 (2-3) 4 (3.8-4) 0.0042
Contrast 2 (1.8-2.3) 2 (1.8-2) 2 (1-2) 0.889
Portal venous phase
Objective image quality
Mean HU portal vein 195.3 ± 21.9 177.7 ± 15.8 239.3 ± 4.9 < 0.0013
Mean HU liver parenchyma 130.6 ± 10.5 121.3 ± 4.9 148.3 ± 6.3 < 0.0014
SNR Liver 8.2 ± 1.1 9.8 ± 1.7 8.6 ± 0.5 0.118
CNR Liver 4.5 ± 1.3 4.5 ± 1.0 4.5 ± 0.4 0.990
Subjective image quality (median, IQR)
Overall 2 (1.8-3) 1.5 (1-2) 2 (2-2.3) 0.0496
Noise 2.5 (2-3) 2 (1-2) 3 (3-3.3) 0.0012
Contrast 2 (1.8-2) 1.5 (1-2) 2 (2-2) 0.222
Lesion detection 1.5 (1-2.3) 1 (1-1.3) 2 (2-2) 0.0372

88

Vol_BM_productie.indd 88 02/02/2022 15:23:03

 Individualized Abdominal CT Protocols

Table 2. Continued

Protocol
Standard CM dose Radiation P
saving dose saving
Delayed phase
Objective image quality
Mean HU liver parenchyma 127.1 ± 7.7 117.1 ± 5.4 142.4 ± 6.3 < 0.0015
SNR Liver 9.2 ± 1.5 9.4 ± 1.2 8.5 ± 1.0 0.504
CNR Liver 4.2 ± 0.8 4.0 ± 0.7 4.5 ± 0.6 0.592
Subjective image quality (median, IQR)
Overall 2.5 (2-3) 2 (1.8-2) 3 (2-3) 0.086
Noise 3 (2-3) 2 (2-2.3) 4 (3.8-4) 0.0022
Contrast 2 (2-2) 2 (1.8-2) 2 (2-2) > 0.99
Lesion detection 2 (2-3) 2 (1.8-2) 2 (2-3) 0.333
Mean attenuation (mean HU), signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR)
in different scan phases using different CM protocols and slider positions, as well as the 5
subjective (overall) image quality. HU indicates Hounsfield Units.
1
Post hoc comparison showed a significant difference between standard and radiation
dose saving (P = 0.01) and between CM dose saving and radiation dose saving (P < 0.001)
2
Post hoc comparison showed a significant difference between CM dose saving and radiation
dose saving.
3
Post hoc comparison showed a significant difference between standard and radiation
dose saving (P = 0.002) and between CM dose saving and radiation dose saving (P < 0.001).
4
Post hoc comparison showed a significant difference between standard and radiation
dose saving (P = 0.005) and between CM dose saving and radiation dose saving (P < 0.001).
5
Post hoc comparison showed a significant difference between standard and CM dose
saving (P = 0.05), between standard and radiation dose saving (P = 0.005) and between CM
dose saving and radiation dose saving (P < 0.001)
6
Post hoc comparison showed no significant difference between groups.

Subjective image quality

Overall subjective image quality and assessment of contrast did not significantly
differ between protocols (table 2). Lesion detection was significantly better in
the CM dose saving protocol compared to the radiation dose saving protocol in
portal venous phase (P = 0.037). The IQR for lesion detection using the standard
protocol varied between good and excellent (1 - 2.3). Using the CM dose saving
protocol the IQR was excellent (1 - 1.3) and using the radiation dose saving IQR
was good (2 – 2). No significant differences in lesion detection were

89

Vol_BM_productie.indd 89 02/02/2022 15:23:03

 Chapter 5

Figure 3. Example of acquired images of repeated scans in a single subject. Three dif-
ferent contrast media (CM) and radiation dose protocols were used. Standard protocol:
350 mgI/kg CM, iodine delivery rate (IDR) 1.1 g I/s , slider position 7; CM dose saving
protocol: 294 mgI/kg CM, IDR 0.9 g I/s, slider position 3; radiation dose saving protocol:
441 mgI/kg CM, IDR 1.3 g I/s, slider position 11.

found in delayed phase (P = 0.333). Noise was rated lowest – corresponding

with a better value on the Likert scale - for the CM dose saving protocol, and
significantly higher for the radiation dose saving protocol (P < 0.01) for all
three phases. Figure 3 shows an example of images acquired from a single pig
scanned several times in arterial and portal venous phases using the three
protocols (standard, CM dose saving, and radiation dose saving).

90

Vol_BM_productie.indd 90 02/02/2022 15:23:03

 Individualized Abdominal CT Protocols

Discussion

The results of the current study show that optimizing either the radiation
or the CM dose is feasible in abdominal CT imaging by combining scan and
injection protocols. Based on an individual risk assessment it seems possible
to reduce either one of the parameters, without negatively influencing the
objective and subjective image quality. Both SNR and CNR were comparable
between groups in all scan phases (arterial, portal venous and delayed phase).
The attenuation of the liver parenchyma was significantly different between
groups in portal venous and delayed phases, however expected based on the
study design. The tube voltage was kept constant in each group (90 kV), while
the CM injection protocol differed between groups. In the radiation dose saving
group TIL was highest and TIL was lowest in the CM dose saving group. The
overall and contrast image quality did not significantly differ between groups.
Noise was rated significantly higher in the radiation dose saving group, in all
scan phases. Lesion detection was good to excellent in portal venous and
5
delayed phase, with a significantly higher score for images acquired in portal
venous phase using the CM dose saving protocol. Overall subjective image
quality was higher for images acquired using the CM dose saving protocol, but
post hoc comparison found no significant difference between groups.

The current study uses a more integrated approach, where previous studies
on this topic have more disconnected set ups (e.g., optimizing CM dose based
on patient body composition or individualizing radiation dose based on ATCM
and ATVS techniques) (2-6, 21-24). The current results show that it is feasible to
adapt either radiation or CM dose to individual risk assessment. As opposed
to a more disconnected approach, using the ATVS slider offers an integrated
concept.

By adjusting the slider settings in the semimode of the ATVS system on a third-
generation dual-source CT scanner, Euler et al. showed that optimizing either
radiation or CM dose led to comparable image quality in low kV CT angiography
imaging, compared to a standard 120 kV exam (13). A 34.3 % reduction in
radiation dose or a 20.2 % reduction in CM dose was feasible without significant
difference in overall subjective image quality among protocols. In vascular
imaging, in general more noise is accepted in order to be able to reliably assess

91

Vol_BM_productie.indd 91 02/02/2022 15:23:03

 Chapter 5

vascular structures because surrounding organs are of less importance. In

parenchymal studies, the balance between noise and attenuation of the organs
is much more delicate. Both excessive noise and insufficient CM attenuation
might result in diagnostic insufficiency, for example an inability to detect liver
lesions. Earlier studies focused on CM reduction in patients with reduced
kidney function. By decreasing tube voltage, a substantial reduction in CM
could be achieved without negatively influencing either objective or subjective
image quality (25, 26). Reducing both parameters at the same time will decrease
CNR, and may lead to insufficient image quality (13). However, in the current
study CNR was comparable between groups by adapting either radiation or
CM dose, as intended in the study design.

Surprisingly, although not significant, the contrast was rated highest in images
acquired using the CM saving protocol for both portal venous and delayed
phases. Possible explanations are twofold. First, even though intra-individual
comparisons provide a unique opportunity for protocol evaluation, the small
population of 6 means that each subjective image quality contributes to a
sixth of the end result. Second, a combination of the factors scored in the
current study (noise, contrast, and lesion detection) determine subjective image
quality, and results may reflect the fact that it is difficult for readers to separate
parameters (27). For example, image quality of a low noise, mediocre contrast
enhancement CT image may still be evaluated ‘good’, because the lack in CM
enhancement is masked by low noise level. Unfortunately, to date, no objective
parameter exists which is able to reliably quantify image quality in a way which
incorporates both objective and subjective aspects.

Image quality depends on both scan parameters (radiation dose related)

and CM injection protocol (CM dose related). Radiation and CM dose can be
calculated for each individual patient and the resulting data manually entered
into scanner and injector devices. ATVS techniques automatically individualize
radiation dose which can be very useful, but the aim is radiation dose reduction
only. Information regarding the CM injection protocol (e.g., CM volume) is not
taken into account despite playing a role in ATVS methodology (11-13). ATCM
and ATVS together optimize radiation dose by adjusting tube current and
tube voltage, based on the clinical question and patient characteristics. By
incorporating the CM injection protocol into these algorithms, protocols can be

92

Vol_BM_productie.indd 92 02/02/2022 15:23:03

 Individualized Abdominal CT Protocols

further adapted to individual requirements, such as for patients with reduced

kidney function or young age, or to specific disease management regimes and
active surveillance.

The current study has some limitations. First, it is a single-center animal study,
and both generalization and translation to humans may be limited. However,
Göttingen minipigs have been shown to be suitable as minipigs are anatomically
comparable to humans (28, 29). Second, as the animals were healthy, no liver
lesions could be evaluated, which makes the parameters ‘lesion detectability’
slightly arbitrary. Another limitation is that the ATVS slider adjustment is a
vendor specific technique and results presented might therefore not directly
be generalizable to other vendors.

In conclusion, in this experimental setup, optimizing either radiation (-26 %) or

CM dose (-16 %) resulted in comparable objective and subjective diagnostic
image quality in abdominal CT. This study demonstrates the feasibility of
5
protocol individualization by adapting a combination of scan and CM injection
parameters, which offers new opportunities for taking into account patient-
related risk factors such as age and kidney function.

93

Vol_BM_productie.indd 93 02/02/2022 15:23:03

 Chapter 5

References

1. Lell MM, Kachelriess M. Recent and upcoming technological developments in

computed tomography: High speed, low dose, deep learning, multienergy. Invest
Radiol. 2020;55(1):8-19.
2. Awai K, Kanematsu M, Kim T, et al. The optimal body size index with which to
determine iodine dose for hepatic dynamic ct: A prospective multicenter study.
Radiology. 2016;278(3):773-81.
3. Kondo H, Kanematsu M, Goshima S, et al. Body size indices to determine iodine
mass with contrast-enhanced multi-detector computed tomography of the upper
abdomen: Does body surface area outperform total body weight or lean body
weight? Eur Radiol. 2013;23(7):1855-61.
4. Heiken JP, Brink JA, McClennan BL, et al. Dynamic incremental ct: Effect of volume
and concentration of contrast material and patient weight on hepatic enhancement.
Radiology. 1995;195(2):353-7.
5. Bae KT, Shah AJ, Shang SS, et al. Aortic and hepatic contrast enhancement with
abdominal 64-mdct in pediatric patients: Effect of body weight and iodine dose.
AJR Am J Roentgenol. 2008;191(5):1589-94.
6. Kondo H, Kanematsu M, Goshima S, et al. Aortic and hepatic enhancement at
multidetector ct: Evaluation of optimal iodine dose determined by lean body weight.
Eur J Radiol. 2011;80(3):e273-7.
7. Martens B, Hendriks BMF, Eijsvoogel NG, et al. Individually body weight-adapted
contrast media application in computed tomography imaging of the liver at 90 kvp.
Invest Radiol. 2019;54(3):177-82.
8. Martens B, Hendriks BMF, Mihl C, Wildberger JE. Tailoring contrast media protocols
to varying tube voltages in vascular and parenchymal ct imaging: The 10-to-10 rule.
Invest Radiol. 2020;55(10):673-6.
9. Martens B, Wildberger JE, Hendriks BMF, et al. A solution for homogeneous liver
enhancement in computed tomography: Results from the complex trial. Invest
Radiol. 2020;55(10):666-72.
10. Martens B, Wildberger JE, Van Kuijk SMJ, et al. Influence of contrast material
temperature on patient comfort and image quality in computed tomography of
the abdomen: A randomized controlled trial. Invest Radiol. 2021.
11. Lurz M, Lell MM, Wuest W, et al. Automated tube voltage selection in
thoracoabdominal computed tomography at high pitch using a third-generation
dual-source scanner: Image quality and radiation dose performance. Invest Radiol.
2015;50(5):352-60.
12. Papadakis AE, Damilakis J. Automatic tube current modulation and tube voltage
selection in pediatric computed tomography: A phantom study on radiation dose
and image quality. Invest Radiol. 2019;54(5):265-72.

94

Vol_BM_productie.indd 94 02/02/2022 15:23:03

 Individualized Abdominal CT Protocols

13. Euler A, Taslimi T, Eberhard M, et al. Computed tomography angiography of the aorta-
optimization of automatic tube voltage selection settings to reduce radiation dose or
contrast medium in a prospective randomized trial. Invest Radiol. 2021;56(5):283-91.
14. The 2007 recommendations of the international commission on radiological
protection. Icrp publication 103. Ann ICRP. 2007;37(2-4):1-332.
15. Board of Radiation Effects Research Division on Earth and Life Sciences National
Research Council of the National Academies. Health risks from exposure to low
levels of ionizing radiation: Beir vii, phase 2. Washington (DC); 2006.
16. Barrett B, Stiles M, Patterson J. Radiation risks: Critical analysis and commentary.
Prev Med. 2012;54(3-4):280-2.
17. Levey AS, de Jong PE, Coresh J, et al. The definition, classification, and prognosis
of chronic kidney disease: A kdigo controversies conference report. Kidney Int.
2011;80(1):17-28.
18. Winklehner A, Goetti R, Baumueller S, et al. Automated attenuation-based tube
potential selection for thoracoabdominal computed tomography angiography:
Improved dose effectiveness. Invest Radiol. 2011;46(12):767-73.
19. Sibulesky L. Normal liver anatomy. Clin Liver Dis (Hoboken). 2013;2(Suppl 1):S1-s3.
20. Jamieson S. Likert scales: How to (ab)use them. Med Educ. 2004;38(12):1217-8. 5
21. Kondo H, Kanematsu M, Goshima S, et al. Body size indexes for optimizing iodine
dose for aortic and hepatic enhancement at multidetector ct: Comparison of total
body weight, lean body weight, and blood volume. Radiology. 2010;254(1):163-9.
22. Matsumoto Y, Masuda T, Sato T, et al. Contrast material injection protocol with
the dose determined according to lean body weight at hepatic dynamic computed
tomography: Comparison among patients with different body mass indices. J
Comput Assist Tomogr. 2019;43(5):736-40.
23. Kaza RK, Platt JF, Goodsitt MM, et al. Emerging techniques for dose optimization in
abdominal ct. Radiographics. 2014;34(1):4-17.
24. Lell MM, Wildberger JE, Alkadhi H, et al. Evolution in computed tomography: The
battle for speed and dose. Invest Radiol. 2015;50(9):629-44.
25. Nagayama Y, Tanoue S, Tsuji A, et al. Application of 80-kvp scan and raw data-based
iterative reconstruction for reduced iodine load abdominal-pelvic ct in patients at
risk of contrast-induced nephropathy referred for oncological assessment: Effects on
radiation dose, image quality and renal function. Br J Radiol. 2018;91(1085):20170632.
26. Sakabe D, Nakaura T, Oda S, et al. Decreasing the radiation dose for contrast-
enhanced abdominal spectral ct with a half contrast dose: A matched-pair
comparison with a 120 kvp protocol. BJR Open. 2020;2(1):20200006.
27. Park HJ, Jung SE, Lee YJ, et al. The relationship between subjective and objective
parameters in ct phantom image evaluation. Korean J Radiol. 2009;10(5):490-5.
28. Siefert J, Hillebrandt KH, Kluge M, et al. Computed tomography-based survey of the
vascular anatomy of the juvenile gottingen minipig. Lab Anim. 2017;51(4):388-96.
29. Takasu M, Tsuji E, Imaeda N, et al. Body and major organ sizes of young mature
microminipigs determined by computed tomography. Lab Anim. 2015;49(1):65-70.
95

Vol_BM_productie.indd 95 02/02/2022 15:23:04

Vol_BM_productie.indd 96 02/02/2022 15:23:05
 CHAPTER
6
Influence of contrast material
temperature on patient
comfort and image quality in
computed tomography of
the abdomen (CATCHY):
a randomized controlled trial

B. Martens, J.E. Wildberger, S.M.J. Van Kuijk, J. De Vos – Geelen, C.R.L.P.N

Jeukens, C. Mihl

Published in:
Investigative Radiology 2022; 57(2):85-89

Vol_BM_productie.indd 97 02/02/2022 15:23:07

 Chapter 6

Abstract

Background: International guideline-recommendations on safe use of contrast

media (CM) are conflicting regarding the necessity to pre-warm iodinated CM.

Purpose: Aim of the study was to evaluate the effects of room temperature CM
compared to pre-warmed CM on image quality, safety and patient comfort in
abdominal computed tomography (CT).

Methods: CATCHY is a double-blinded, randomized non-inferiority trial. Between

February and August 2020, 218 participants referred for portal venous
abdominal CT were prospectively and randomly assigned to one of two groups.
All patients received iopromide at 300mgI/ml: group 1 at room temperature
(~23°C [~73°F]), group 2 pre-warmed to body temperature (37°C [99°F]). A state-
of-the-art individualized CM injection protocol was used, based on body weight
and adapted to tube voltage. Primary outcome was absolute difference in mean
liver attenuation between groups, calculated with a two-sided 95% confidence
interval. The non-inferiority margin was set at -10HU. Secondary outcomes
were objective (signal-to-noise [SNR] and contrast-to-noise ratios [CNR]) and
subjective image quality; CM extravasations and other adverse events; and
participant comfort (five-point scale questionnaire) and pain (numeric rating
scale). This trial is registered with ClinicalTrials.gov (NCT04249479).

Results: The absolute difference in mean attenuation between groups was

+4.23HU (95% CI +0.35 to +8.11; mean attenuation 122.2±13.1HU in group 1,
118.0±15.9HU in group 2; P=0.03). SNR, CNR and subjective image quality were
not significantly different between groups (P=0.53, 0.23 and 0.99 respectively).
Contrast extravasation occurred in one patient (group 2), no other adverse
events occurred. Comfort scores were significantly higher in group 1 than in
group 2 (P=0.03), pain did not significantly differ (perceived P>0.99; intensity
P=0.20).

Conclusion: Not pre-warming iodinated CM was found non-inferior in abdominal

CT imaging. Pre-warming conferred no beneficial effect on image quality, safety,
and comfort, and might therefore no longer be considered a prerequisite in
state-of-the art injection protocols for parenchymal imaging.

98

Vol_BM_productie.indd 98 02/02/2022 15:23:07

 CM Temperature and Patient Comfort in Abdominal CT

Key words

Multidetector Computed Tomography; Diagnostic Imaging; Abdomen; Contrast

Media; Image Quality; Contrast Material Warming; Discomfort.

Introduction

Computed tomography (CT) has rapidly evolved (1-3). Both scan and contrast
media (CM) protocols have been individualized based on patient characteristics
as well as for clinical indications (4-8). The effects of various characteristics of
CM have been thoroughly investigated (8-16). Among these, CM viscosity is
key. In general, viscosity of CM increases with higher CM concentration and
is directly influenced by temperature: pre-warming CM leads to decreased
viscosity, which may reduce the risk of both CM extravasation and adverse
events in general, increasing participant comfort (14, 15, 17). However, the
necessity to pre-warm CM for clinical CT applications is still under debate (8-
13). Indeed, European and American guidelines on the use of CM are not in
6
agreement on pre-warming CM (12, 13). The European Society of Urogenital
Radiology recommends pre-warming iodine-based CM in all cases (13). On the
other hand, according to the American College of Radiology, pre-warming CM
is only necessary for concentrations of 370mg iodine per ml or higher, injection
rates above 5ml/s, or if small-gauge catheters are used (12, 18, 19). The latter
advice is primarily based on a large retrospective study by Davenport et al.
comparing 12.682 injections with pre-warmed CM to 12.138 injections without
pre-warmed CM (11). Adverse event rates were not different for iopamidol 300
injections of less than 6ml/s, but were significantly reduced by pre-warming
for iopamidol 370 injections.

Pre-warming CM requires special equipment and more complex planning and

logistics. On the other hand, pre-warming CM may yield higher attenuation
levels, image quality and comfort (20). The question remains whether pre-
warming CM is necessary when moderate flow rates (<6ml/s) are used, as is
the case in abdominal imaging.

The aim of the study ContrAst media Temperature and patient Comfort in
computed tomograpHY of the abdomen (CATCHY), was to prospectively

99

Vol_BM_productie.indd 99 02/02/2022 15:23:07

 Chapter 6

compare room temperature CM to pre-warmed CM with regard to image

quality, safety, and participant comfort in portal venous abdominal imaging.

Materials and Methods

Ethics

This double-blinded randomized controlled non-inferiority trial was approved

by the local ethics committee and the institutional review board, and is
registered on ClinicalTrials.gov (NCT04249479). Written informed consent was
obtained before inclusion in the clinical trial. The study did not receive any
industry support.

Study design & Study Population

Using CM at room temperature (~23°C [~73°F]) might result in lower attenuation

than would be achieved using CM pre-warmed to body temperature (20). The
hypothesis of the CATCHY-trial is that using CM at room temperature does not
compromise diagnostic image quality, patient safety or comfort in the setting
of abdominal imaging. The sample size was calculated to enable detection of an
absolute difference of 10HU in mean attenuation of the liver. This non-inferiority
margin was chosen based on earlier studies where mean attenuation of 120HU
was found sufficient and a decrease in attenuation of 10% was pronounced
clinically significant (21). To be able to detect a difference greater than 10HU
with a power of 90% and two-sided alpha of 5%, 98 participants per group
are required. We recruited an additional 10% to account for potential loss to
follow-up.

Participants referred for an abdominal CT in portal venous phase, were

prospectively included between February and August 2020 at our center.
Exclusion criteria were hemodynamic instability, pregnancy, renal insufficiency
(estimated glomerular filtration rate <30 mL/min per 1.73m2), prior adverse
reactions to iodinated CM, age below 18 and inability to place an 18-gauge
needle (22, 23). Additional scanning was no reason for exclusion unless it
altered the underlying CM injection protocol. Repeated inclusion was allowed,
as it was not expected to influence outcome. Body weight (kg) of the participant

100

Vol_BM_productie.indd 100 02/02/2022 15:23:07

 CM Temperature and Patient Comfort in Abdominal CT

was measured prior to scanning on a calibrated scale. As the maximum level

of the dual head injector syringes is 200ml, participants with a body weight
>115kg were excluded from this study. Participants’ height (meters) was checked
and Body Mass Index calculated. A 1:1 computer-generated randomization
schedule was used (TENALEA, Trans European Network for Clinical Trials
Services). Stratification factors were age (<60 and ≥60years) and weight (<75
and ≥75kg). Participants were equally divided in two groups by variable block
randomization.

All data was collected by one blinded researcher (B.M.) using electronic case-
report forms and checked by an independent study monitor. Patients were
blinded as to the allocated treatment. A written questionnaire evaluating
comfort was filled in by the participant directly after each CT exam.

Scan and contrast media injection protocol

A third-generation dual source CT scanner (Somatom Force; Siemens

6
Healthineers, Forchheim, Germany) with automated tube current modulation
(CareDose4D, Siemens) and automated tube voltage selection (CarekV; Siemens)
techniques was used: 120kVref and 150mAsref, 192x0.6 mm slice collimation,
gantry rotation time 0.5s.

Group 1 received CM at room temperature (~23°C [~73.4°F]) and group 2

received pre-warmed CM (37°C [99°F]). An 18-gauge catheter (Vasofix®
Safety, B Braun, Melsungen, Germany) was placed by the radiographer in the
participants’ arm (e.g. antecubital vein, forearm or wrist) prior to scanning. For
both groups, the CM injection protocol (300mg/ml [Iopromide; Ultravist 300;
Bayer Healthcare, Berlin, Germany]) was adapted to the participants’ body
weight and the tube voltage used (at a tube voltage of 120, 110, 100 and 90kV
a respective dosing factor of 0.521, 0.469, 0.417 and 0.365gI/kg was used) (8,
21). The scan was performed 70s after start of the CM injection using a dual
head power injector (Stellant®; Bayer Healthcare, Berlin, Germany) followed
by a saline flush with the same injection speed and an overall volume of 40ml.

101

Vol_BM_productie.indd 101 02/02/2022 15:23:07

 Chapter 6

Image reconstruction parameters were: 3mm slice thickness, 2mm increment,

soft tissue kernel (Br40) and iterative reconstruction (IR) strengths 2/3 (Siemens;
Advanced Modelled Iterative Reconstruction).

A dedicated data acquisition program (Certegra Informatics Solution; Bayer)

monitored the CM parameters. Radiation dose and reconstruction settings
were collected from the dose sheet at the PACS workstation (IMPAX version
6.6.1.5003; AGFA HealthCare N.V., Mortsel, Belgium).

Primary outcome

Absolute difference in mean attenuation of the liver parenchyma between

groups was calculated with a two-sided 95% confidence interval (CI) of the
difference. Mean attenuation in Hounsfield Units (HU) was based on three
liver segments, preferably segments 2, 5 and 8 (Couinaud classification (24)).
A region of interest was drawn in each segment (area: ≥1cm2), not containing
vessels, bile ducts or lesions.

Secondary outcomes

Objective image quality was rated using signal-to-noise ratio (SNR: mean
attenuation divided by the mean standard deviation [SD]) and contrast-to-
noise ratio (CNR: mean liver attenuation minus HU of the left paraspinal muscle,
divided by the SD of the attenuation of the paraspinal muscle). Subjective image
quality was rated in consensus on a 5-point Likert scale by two readers, B.M.
and C.M. (5- and 10-years’ experience in abdominal imaging, respectively).
Readers were blinded to the allocated protocol. Overall image quality was rated
on a 5-point Likert scale (1=excellent, 2=good, 3=moderate, 4=poor and 5=very
poor) (21). Readers were allowed to level window settings individually.

All adverse events, including contrast extravasation, were reported by the

radiographer.

Comfort was rated by the patient on the questionnaire provided by the

radiographer directly after the scan was performed (1=very bad, 2=bad,
3=neutral, 4=good, 5=excellent). An 11-point numeric rating scale was used to

102

Vol_BM_productie.indd 102 02/02/2022 15:23:07

 CM Temperature and Patient Comfort in Abdominal CT

evaluate pain during injection (0=no pain; 10=very severe pain) (17). Feelings of
shivering, goosebumps or cold were evaluated and an open field was provided
for the patient to record any other experiences. The questionnaire is given in
the Appendix.

Statistical analysis

Dichotomous outcomes are reported as absolute numbers with percentages,

continuous outcome variables are reported as means ±SD. Results are stratified
by treatment allocation. To test for non-inferiority, a CI approach was used
on an analysis of covariance (ANCOVA) model, with a two-sided 5% level of
significance. For the primary endpoint, non-inferiority of room temperature CM
to pre-warmed CM could be claimed if the lower limit of the CI for the absolute
difference in mean liver attenuation (room temperature CM group minus pre-
warmed CM group) falls above -10HU. This test for non-inferiority was only
performed for the primary outcome.

6
Figure 1. Trial profile. Abbreviations: GFR, glomerular filtration rate.

Participant comfort and pain intensity were compared between groups using
the Mann-Whitney U test. The χ2 test, and in case of expected cell counts

103

Vol_BM_productie.indd 103 02/02/2022 15:23:07

 Chapter 6

of less than five, Fishers exact test, was used for dichotomized variables.
Continuous normally distributed variables were compared between groups
using the independent samples t-test. The Mann-Whitney U test was used for
not normally distributed variables. Data was analysed using statistical software
(SPSS, version 26.0; IBM Corp., New York, NY). A two-sided P value <0.05 is
considered statistically significant. Both per-protocol and intention-to-treat
analyses were done.

Table 1. Key demographic and clinical characteristics of randomized groups.

Characteristics Group 1 (Room temperature) Group 2 (37°C [99°F])

N = 109 N = 113
Excluded participants 0 4
Age (y) 66.3 ± 10.6 65.1 ± 11.1
Sex (% male) 56.0% 64.2%
Body weight (kg) 79.7 ± 13.7 78.4 ± 12.8
Height (m) 1.72 ± 0.09 1.72 ± 0.08
BMI (kg/m2) 26.8 ± 3.7 26.5 ± 4.0
Scan indication (%)
Oncology 97.2% 92.7%
Infection 0.9% 2.8%
Other 1.8% 4.6%
Abbreviations: BMI, Body mass index.

Results

Baseline characteristics

Two-hundred twenty-two participants were enrolled. Four participants were

excluded; two due to technical problems, in one participant the questionnaire
form was missing, and one participant already had a 20-gauge needle in place
and therefore had to be excluded (figure 1). All patients received their allocated
treatment. Therefore, in this study, the intention-to-treat population is the same
as the per-protocol population. Key demographic and clinical characteristics
are detailed in table 1.

104

Vol_BM_productie.indd 104 02/02/2022 15:23:07

 CM Temperature and Patient Comfort in Abdominal CT

Table 2. Contrast media and radiation dose parameters.

CM and radiation dose Group 1 (Room Group 2 (37° C) P

parameters temperature) N = 109
N = 109
CM volume (ml) 103.6 ± 21.7 100.8 ± 21.4 0.33
TIL (g) 31.1 ± 6.5 30.2 ± 6.4 0.34
Flow rate (ml/s) 3.4 ± 0.7 3.3 ± 0.7 0.31
Peak flow rate (ml/s) 3.8 ± 0.9 3.8 ± 0.8 0.89
Peak Pressure (psi) 63.1 ± 19.7 54.9 ± 18.4 0.001
IDR (gI/s) 1.0 ± 0.2 1.0 ± 0.2 0.33
Tube voltage (kV) 0.84
90kV (%) 64.2% 67.9%
100kV (%) 33.9% 30.3%
110kV (%) 1.8% 1.8%
120kV (%) 0% 0%
Mean mAsref 293.4 ± 55.9 300.7 ± 57.2 0.29
Mean mAseff 217.1 ± 51.6 208.4 ± 47.0 0.20
CTDIvol (mGy) 7.2 ± 2.3 6.9 ± 2.0 0.31 6
DLP (mGy*cm) 386.5 ± 133.6 390.4 ± 136.8 0.83
IR2(%) / IR3(%) 60.6% / 39.4% 54.1% / 45.9% 0.34
Abbreviations: CM, contrast media; TIL, total iodine load; IDR, iodine delivery rate; mAsref,
quality reference mAs; mAseff, effective mAs, CTDIvol, CT dose index; DLP, dose length product;
IR, Iterative reconstruction. CM volume, TIL, mAseff and DLP, mean liver attenuation (HU), SNR
and CNR were normally distributed and compared using the independent samples T-test.
Flow rate, peak flow rate, peak pressure, mAsref and CTDIvol were not normally distributed
and the Mann-Whitney U test was used.

Mean CM volume was 103.6±21.7ml in group 1 and 100.8±21.4ml in group 2

(P=0.33). Mean flow rate was 3.4±0.7 in group 1 and 3.3±0.7ml/s in group 2
(P=0.31). Other CM and radiation dose parameters are shown in Table 2. There
were no significant differences between groups, except peak pressure (in psi)
which was significantly higher in group 1: 63.1±19.7psi (room temperature CM)
versus 54.9±18.4psi (pre-warmed CM) (P=0.001).

Primary outcome

The percentage difference in mean attenuation (group 1 minus group 2) was

+4.23HU with 95% CI +0.35 to +8.11. The lower limit of the CI of the difference

105

Vol_BM_productie.indd 105 02/02/2022 15:23:07

 Chapter 6

falls within the non-inferiority margin, indicating non-inferiority of room

temperature CM with respect to attenuation (figure 2).

Secondary outcomes

Objective and subjective image quality results are shown in table 3. Mean
attenuation was 122.2±13.1 in group 1 and 118.0±15.9 in group 2 (p=0.03). SNR,
CNR and subjective image quality did not significantly differ between groups
(P=0.53, 0.23 and 0.99 respectively).

There was one person with a contrast extravasation in group 2. No other

adverse events were reported.

Figure 2. Absolute difference in mean attenuation of the liver (room temperature CM

group minus pre-warmed CM group). The dashed line shows the non-inferiority margin,
set at -10HU. Error bars indicate the 95% confidence interval (CI) of the difference; the
bullet shows the point estimate. Abbreviations: HU, Hounsfield Units.

Patient comfort and pain results are shown in table 4. Comfort scores were
higher in group 1 than in group 2 (P=0.03). Comfort was graded excellent or
good by 91.7% of the participants in group 1 and by 86.2% of the participants
in group 2. Comfort was rated bad or very bad by 1 participant (0.9%) in each
group. In group 1 three patients (3.3%) and in group 2 four patients (4.4%)
perceived pain (P>0.99). Pain intensity scores were not significantly different
between groups (P=0.20). Four participants had a feeling of being cold, of which
three were randomized in group 1 (table 4).

106

Vol_BM_productie.indd 106 02/02/2022 15:23:07

 CM Temperature and Patient Comfort in Abdominal CT

Table 3. Objective and subjective image quality.

Group 1 (room Group 2 (37°C P

temperature) [99°F])
N = 109 N = 109
Objective image quality
Mean Attenuation (HU) 122.2 ± 13.1 118.0 ± 15.9 0.03
SNR 9.8 ± 2.1 9.6 ± 2.1 0.53
CNR 6.2 ± 2.4 5.8 ± 2.2 0.23
Subjective image quality
Excellent (%) 26.6% 25.7% 0.99
Good (%) 66.1% 66.1%
Moderate (%) 6.4% 7.3%
Poor (%) 0.9% 0.9%
Very poor (%) 0% 0%
Abbreviations: HU, Hounsfield Units; SNR, signal-to-noise ratio; CNR, contrast-to-noise ratio.

Table 4. Participant comfort.

6
Comfort Group 1 (Room Group 2 (37°C [99°F]) P
temperature) N = 109
N = 109
Contrast extravasation 0 1
Comfort (median, IQR) 4 (4-5) 4 (4-5) 0.03
Excellent (N) 54 39
Good (N) 46 55
Neutral (N) 8 14
Bad (N) 1 0
Very bad (N) 0 1
Pain intensity (median, 0 (0-0) 0 (0-0) 0.20
IQR)
Pain (yes/no) 3/106 4/105 >0.99
Feeling cold
Shivering (N) 0 0
Goosebumps (N) 0 0
Cold (N) 3 1
Abbreviations: IQR, interquartile range.

107

Vol_BM_productie.indd 107 02/02/2022 15:23:07

 Chapter 6

Discussion

CM at room temperature was found to be non-inferior to pre-warmed CM

in mean attenuation of the liver. Furthermore, the present study found no
evidence or benefits from pre-warming iodinated CM with regard to image
quality, safety and patient comfort in portal venous abdominal CT imaging.
Mean attenuation was significantly higher in the room temperature CM group.
Differences in SNR, CNR and subjective image quality between groups were
small and non-significant. Injecting CM at room temperature did not result in
CM extravasations or other adverse events at the given IDR of 1.0gI/s, which
is in line with the results of the study by Davenport et al. (11-13). CM at room
temperature yielded significantly higher participant comfort scores, although
absolute differences are small and may not be clinically relevant.

This is the first prospective randomized trial providing high level evidence that
participant comfort and image quality are not increased by pre-warming CM in
this setting. The European and American guidelines have a conflicting opinion
on this subject (12, 13). Based on the results of the current study it appears
that the American College of Radiology guidelines is the one to follow. CM
extravasation, other adverse events and participant comfort are not adversely
affected by administering CM at room temperature. As a consequence, one may
forego pre-warming for CM injections with low iodine concentration of 300mg/
ml, at moderate flow rates and a catheter of 18-gauge.

Peak pressure was significantly higher for CM at room temperature compared

to pre-warmed CM. Mean flow rate in present study was quite low (mean flow
rate of 3.4ml/s ranging from 2.0ml/s up to a maximum of 5.3ml/s). A higher
flow rate is expected to further increase peak pressure and therefore might
negatively influence participant comfort. However, at our center more than
90% of the scans performed between 2013 and 2019 had a flow rate below
6ml/s with rather low psi and IDR values, and the results of the current trial will
apply. Accordingly, future research could focus on participant comfort when
CM is injected at room temperature with higher flow rates, for example in
cardiovascular imaging. As shown by Davenport et al. increasing flow rates
increases incidences of CM extravasation and other adverse events in specific
settings, most likely also decreasing participant comfort (11).

108

Vol_BM_productie.indd 108 02/02/2022 15:23:07

 CM Temperature and Patient Comfort in Abdominal CT

The study has some limitations. First, it is a single-center randomized controlled

trial. Generalizability to other centers might be limited. Intra-individual
comparison may have been preferable but is not readily feasible in clinical
setting. Second, the sample size was based on a non-inferiority margin for
objective image quality, because not much is known about patient comfort
margins, which was the main study outcome. In addition, CM temperature
was measured in the bottle. Pre-warmed CM might cool down when travelling
through the tubing from the bottle to the patient. Therefore, injected CM
temperature may have been overestimated. Lastly, the mean flow rate was
quite low and results might have been different if a CM injection protocols with
higher flow rate were used.

Pre-warming CM is not beneficial in terms of image quality, safety, and

participant comfort in portal venous phase abdominal CT imaging. Pre-warming
CM should therefore not be a pre-requisite in state-of-the art injection protocols
for parenchymal imaging for CM injections with low iodine concentrations, at
moderate flow rates and a reasonable catheter size.
6

109

Vol_BM_productie.indd 109 02/02/2022 15:23:07

 Chapter 6

References

1. Lell MM, Wildberger JE, Alkadhi H, et al. Evolution in computed tomography: The
battle for speed and dose. Invest Radiol. 2015;50(9):629-44.
2. Lell MM, Kachelriess M. Recent and upcoming technological developments in
computed tomography: High speed, low dose, deep learning, multienergy. Invest
Radiol. 2020;55(1):8-19.
3. Kwan AC, Pourmorteza A, Stutman D, et al. Next-generation hardware advances in
ct: Cardiac applications. Radiology. 2021;298(1):3-17.
4. Kondo H, Kanematsu M, Goshima S, et al. Body size indices to determine iodine
mass with contrast-enhanced multi-detector computed tomography of the upper
abdomen: Does body surface area outperform total body weight or lean body
weight? Eur Radiol. 2013;23(7):1855-61.
5. Awai K, Kanematsu M, Kim T, et al. The optimal body size index with which to
determine iodine dose for hepatic dynamic ct: A prospective multicenter study.
Radiology. 2016;278(3):773-81.
6. Fleischmann U, Pietsch H, Korporaal JG, et al. Impact of contrast media concentration
on low-kilovolt computed tomography angiography: A systematic preclinical
approach. Invest Radiol. 2018;53(5):264-70.
7. Nakaura T, Nakamura S, Maruyama N, et al. Low contrast agent and radiation dose
protocol for hepatic dynamic ct of thin adults at 256-detector row ct: Effect of
low tube voltage and hybrid iterative reconstruction algorithm on image quality.
Radiology. 2012;264(2):445-54.
8. Martens B, Hendriks BMF, Mihl C, Wildberger JE. Tailoring contrast media protocols
to varying tube voltages in vascular and parenchymal ct imaging: The 10-to-10 rule.
Invest Radiol. 2020;55(10):673-6.
9. Bae KT. Intravenous contrast medium administration and scan timing at ct:
Considerations and approaches. Radiology. 2010;256(1):32-61.
10. Mihl C, Wildberger JE, Jurencak T, et al. Intravascular enhancement with identical
iodine delivery rate using different iodine contrast media in a circulation phantom.
Invest Radiol. 2013;48(11):813-8.
11. Davenport MS, Wang CL, Bashir MR, et al. Rate of contrast material extravasations
and allergic-like reactions: Effect of extrinsic warming of low-osmolality iodinated
ct contrast material to 37 degrees c. Radiology. 2012;262(2):475-84.
12. American College of Radiology 2021;Pages. Accessed at https://www.acr.org/-/media/
ACR/Files/Clinical-Resources/Contrast_Media.pdf. Accessed March 4th 2021.
13. European Society of Urogenital Radiology 2018;Pages. Accessed at http://www.esur.
org/fileadmin/content/2019/ESUR_Guidelines_10.0_Final_Version.pdf. Accessed
March 4th 2021.

110

Vol_BM_productie.indd 110 02/02/2022 15:23:08

 CM Temperature and Patient Comfort in Abdominal CT

14. Kok M, Mihl C, Mingels AA, et al. Influence of contrast media viscosity and temperature
on injection pressure in computed tomographic angiography: A phantom study.
Invest Radiol. 2014;49(4):217-23.
15. Roth R, Akin M, Deligonul U, Kern MJ. Influence of radiographic contrast media
viscosity to flow through coronary angiographic catheters. Cathet Cardiovasc Diagn.
1991;22(4):290-4.
16. Zopfs D, Reimer RP, Sonnabend K, et al. Intraindividual consistency of iodine
concentration in dual-energy computed tomography of the chest and abdomen.
Invest Radiol. 2021;56(3):181-7.
17. Kok M, Mihl C, Hendriks BM, et al. Patient comfort during contrast media
injection in coronary computed tomographic angiography using varying contrast
media concentrations and flow rates: Results from the eicar trial. Invest Radiol.
2016;51(12):810-5.
18. Turner E, Kentor P, Melamed JL, et al. Frequency of anaphylactoid reactions
during intravenous urography with radiographic contrast media at two different
temperatures. Radiology. 1982;143(2):327-9.
19. Vergara M, Seguel S. Adverse reactions to contrast media in ct: Effects of
temperature and ionic property. Radiology. 1996;199(2):363-6.
20. Hazirolan T, Turkbey B, Akpinar E, et al. The impact of warmed intravenous contrast
material on the bolus geometry of coronary ct angiography applications. Korean J 6
Radiol. 2009;10(2):150-5.
21. Martens B, Wildberger JE, Hendriks BMF, et al. A solution for homogeneous liver
enhancement in computed tomography: Results from the complex trial. Invest
Radiol. 2020;55(10):666-72.
22. Davenport MS, Perazella MA, Yee J, et al. Use of intravenous iodinated contrast
media in patients with kidney disease: Consensus statements from the american
college of radiology and the national kidney foundation. Radiology. 2020;294(3):660-
8.
23. Nijssen EC, Rennenberg RJ, Nelemans PJ, et al. Prophylactic hydration to protect
renal function from intravascular iodinated contrast material in patients at high risk
of contrast-induced nephropathy (amacing): A prospective, randomised, phase 3,
controlled, open-label, non-inferiority trial. Lancet. 2017;389(10076):1312-22.
24. Germain T, Favelier S, Cercueil JP, et al. Liver segmentation: Practical tips. Diagn Interv
Imaging. 2014;95(11):1003-16.

111

Vol_BM_productie.indd 111 02/02/2022 15:23:08

 Chapter 6

Appendix

Patient comfort assessment questionnaire. Note that this is a translation, as

the original questionnaire was in the native language (Dutch).

112

Vol_BM_productie.indd 112 02/02/2022 15:23:08

 CM Temperature and Patient Comfort in Abdominal CT

113

Vol_BM_productie.indd 113 02/02/2022 15:23:08

Vol_BM_productie.indd 114 02/02/2022 15:23:09
 CHAPTER
7
Tailoring contrast media protocols to
varying tube voltages in vascular and
parenchymal CT imaging

The 10-to-10 rule

B. Martens*, B.M.F. Hendriks*, C. Mihl, J.E. Wildberger

Published in:
Investigative Radiology 2020; 55(10):673-676 (* shared first authorship)

Vol_BM_productie.indd 115 02/02/2022 15:23:10

 Chapter 7

Abstract

The latest technical developments in CT have created the possibility for

individualized scan protocols, using variable kV settings. A lower tube voltage
draws closer to the iodine K-edge, increasing iodine attenuation, compared
to a higher tube voltage. In addition, attenuation is influenced by patient
characteristics such as total body weight. Therefore, in order to maintain a
robust contrast enhancement throughout the patient population in both
vascular and parenchymal CT scans, one must adapt the contrast media (CM)
protocol to the kV setting and the patient body habitus. This paper presents
current evidence and proposes a simple rule-of-thumb for adjusting the CM
protocol in both vascular and parenchymal studies; the 10-to-10 rule.

Key words

Computed tomography; Contrast media; Radiation Dosage; Abdomen;

Pulmonary artery; Coronary vessels; Aorta.

116

Vol_BM_productie.indd 116 02/02/2022 15:23:10

 Contrast Media Protocols: The 10-to-10 Rule

Introduction

Since the development of the first computed tomography (CT) scanner in 1967,
the technology of these scanners has evolved (1). The advent of powerful x-ray
tubes, multiple detector rows and dual source technology has led to short
acquisition times and excellent temporal and spatial resolution (2).

CT nowadays is a widely available, versatile and fast medical imaging method

which has revolutionized radiology and the field of medicine as a whole (3).
In the era of personalized medicine, we are increasingly deviating from the
“one size fits all protocol” and moving towards scan and contrast media (CM)
injection protocols tailored to the individual patient.

The paper summarizes the most relevant factors in CM protocols for optimal
attenuation in parenchymal CT and CT angiography (CTA). It proposes an easy-
to-use rule of thumb (the 10-to-10 rule) for tailoring CM injection protocols
to variable kV settings (4-6). When used in conjunction with personalized CM
protocols, a homogeneous image quality throughout the patient population
can be achieved. 7
Individualized scan protocols

A comprehensive overview of the latest technical developments in CT has

recently been published by Lell and colleagues (3). The latest tube technology
allows for higher ion flux during longer acquisition times; this translates in
clinical practice to lower kV scanning and lower radiation dose for most patients
(2). Automated tube current modulation (ATCM) is a technique used often
in daily clinical practice and recently the automated tube voltage selection
(ATVS) technique was introduced. These advances in CT technology create the
opportunity for radiation dose tailoring to each individual patient and scan
indication (7, 8). Where a large adult might be scanned at 120 kV in order to
achieve good image quality, a smaller person may be scanned using a tube
voltage as low as 70 or 80 kV and concordantly the tube current is adjusted
in order to achieve similar image quality with a subsequent lower radiation
dose. The possibility for low kV scanning is advantageous for lowering the

117

Vol_BM_productie.indd 117 02/02/2022 15:23:10

 Chapter 7

hypothetical radiation-induced cancer risk for patients and follows the “As Low
As Reasonable Achievable” (ALARA) principle (9).

Individualized contrast media protocols

In parenchymal studies the total volume is most important for reaching optimal
liver enhancement, while previous studies have indicated that the iodine
delivery rate (IDR, in g I/s) is the decisive factor for determining intravascular
enhancement in vascular CT (10). This concept has been proven for CTA of
the pulmonary arteries (CTPA), coronary CTA (CCTA) and CTA of the aorta (11-
14). The relationship between IDR, flow rate and CM concentration can be
explained with the following formula: IDR (g I/s) = [CM] (g I/ml) * flow rate (ml/s).
Normalizing IDR is a straightforward approach to make different injection
protocols comparable. As seen in the formula, IDR can be modified either by
adapting flow rate or adapting CM concentration.

Several patient factors are known to influence the attenuation of both vascular
and parenchymal structures. Patient body weight is a well described influential
factor in CM application (15). In coronary and pulmonary arteries and the liver
parenchyma, adapting the CM volume to the patients’ weight has proven to
be beneficial (see figure 1) (13, 16, 17). Adapting the volume to the total body
weight (TBW) results in a more homogeneous attenuation of pulmonary and
coronary arteries and the liver parenchyma, between patients,

in comparison to a fixed ‘one size fits all’ CM injection protocol. An individualized

CM injection protocol also turned out to reduce CM usage in general, while
reaching optimal enhancement levels (13, 16, 17).

It is possible to individualize CM injection protocols based on TBW as mentioned

above, but factors as lean body weight (LBW), body mass index (BMI) and body
surface area (BSA) are also parameters widely studied (18). LBW is TBW minus
the amount of fat the body contains, with the underlying idea that fat is less
vascularized compared to muscle (10). Therefore, considering this concept,
fat should not be taken in to account when calculating the optimal amount of
CM that should be given to a specific patient. LBW can be calculated by using

118

Vol_BM_productie.indd 118 02/02/2022 15:23:10

 Contrast Media Protocols: The 10-to-10 Rule

either the James or the Boer formula, of which the latter is the first choice in
obese patients (19).

Figure 1. Individualized contrast media injection protocols based on body weight, com-
pared to a fixed injection protocol in the pulmonary arteries, the proximal coronary
arteries and in the liver parenchyma. The figure depicts more similar and robust en-
hancement of both vascular and parenchymal structures with individualized protocol
(modified from: 10, 13, 14 and presented schematically).

Fixed protocol Individualized protocol

Pulmonary Arteries

Coronary Arteries

Liver Parenchyma
Hounsfield Units

Total body weight Total body weight 7

Combining scan- and contrast media protocols for optimal image quality

With lower tube voltages and therefore the x-ray output drawing closer to
the 33 keV k-edge of iodine, the photoelectric effect increases which in turn
increases the attenuation of iodine (10). This provides new opportunities for
CM individualization in both arterial and parenchymal studies. The benefits
of lower kV scanning are twofold. First of all, it allows for the possibility of
reducing the total amount of CM, hypothesized to be beneficial in preventing
contrast-induced nephropathy (20, 21). Some controversy remains on whether
intravenous application of CM causes the sometimes observed and reversible
dip in renal function (22, 23). Nevertheless, there is simply no need to give
patients more CM than needed, especially as the underlying physiological
effects are still not fully understood.

The use of variable, individualized kV-settings with the arrival of ATCM and
ATVS techniques, comes with a new challenge; the variety in kV settings used

119

Vol_BM_productie.indd 119 02/02/2022 15:23:11

 Chapter 7

have a substantial impact on the attenuation of iodine. If the iodine k-edge

effect is not taken into account, the CT attenuation numbers and thereby image
quality in a patient population scanned with different tube voltage settings is
heterogeneous (11). Radiologists incorporate attenuation characteristics when
drawing conclusions from CT images. For example, the assessment of lesions
to be more likely benign or malignant depends on the attenuation pattern
(24, 25). With a large variety of tube voltages used and a great diversity in CM
volumes and patient characteristics in daily clinical practice, it is important to
reach a comparable enhancement of the target structure, regardless of these
factors. Only when attenuation is similar and robust between patients, reliable
conclusions can be drawn from radiological imaging.

Several groups have investigated the effect of low kV scanning on attenuation

values during CTA and parenchymal CT (16, 18, 26-28). At present, one important
enigma remains; how to correctly apply patient-specific, tube voltage tailored
CM protocols for each clinical indication and scan technique. The possibilities
of lowering the kV setting and the methods of concordantly adapting the CM
parameters have been studied widely. The same accounts for the patient-
specific individualisation of CM protocols. The authors combined the various
findings of these studies into a practical, easy to remember, rule of thumb: the
10 to 10 rule.

Rule of thumb: The 10 to 10 rule

Several IDR reduction percentages have been proposed when trying to adapt
the vascular CT protocol to the different tube voltages used and some have
been validated in a clinical setting (4, 28, 29). When looking closely at the
available literature there is some overlap in previously described methods,
which can be boiled down to a rule-of-thumb: In CTA one may roughly deduct
10 % of the IDR per 10 kV step down and vice versa. A straightforward way to
adapt the IDR is by changing the flow rate, however can also be achieved by
altering the concentration of iodine in the CM. This rule has been validated in
clinical practice for CTPA and CCTA (4, 5).

In parenchymal studies a reduction of 10 kV should result in a 10 % reduction in

total iodine load (TIL) (30). The landmark paper by Heiken et al. from 1995 stated

120

Vol_BM_productie.indd 120 02/02/2022 15:23:11

 Contrast Media Protocols: The 10-to-10 Rule

that an attenuation increase of ≥ 50 HU of the liver parenchyma is necessary to

ensure appropriate visibility of low-attenuating lesions, based on a tube voltage
of 120 kV (31, 32). To achieve this attenuation, a dosing factor of 0.521 grams of
iodine per kilogram (g I/kg) can be calculated (31). A recent study showed that
a 10 kV reduction should lead to a 10 % decrease of the dosing factor and vice
versa, to be able to individualize the parenchymal CT based on both patient
and scanner characteristics (6, 33).

This 10-to-10 rule can easily be applied to any existing patient tailored protocol,
thereby adjusting the CM protocol for any kV setting and individual patient
characteristics (see figure 2).

Figure 2. The rule of thumb.

Example I: Vascular studies

When a patient is scanned with a reduced tube voltage of 70 kV, where

normally 120 kV is used, one may decrease the IDR by 50 % (34). When the
same concentration of CM is used, this could mean a 50 % reduction in flow
rate. A 50 % decrease in flow rate at a constant injection time will also lead to
a reduction in total CM volume and thereby the TIL received by the patient.
In addition, keeping the injection time constant ensures it to be a robust and
reliable method to be used 24/7.

Figure 3 is an example of the same patient scanned twice with different kV

settings and adapted IDR, which resulted in a similar enhancement of the
pulmonary arteries (4).

121

Vol_BM_productie.indd 121 02/02/2022 15:23:11

 Chapter 7

Figure 3. Two CT pulmonary angiography scans performed in the same patient be-
cause of possible pulmonary embolism. The iodine delivery rate (IDR) is adapted to the
different kV setting used in each scan; 80 kV with an IDR of 1.02 g I/s for scan A and 70
kV with an IDR of 0.84 g I/s for scan B. The region of interest (circle) in each scan shows
the comparable Hounsfield Units per scan despite the different kV setting used.

There is more than 10 % reduction in IDR, because the patient lost weight
between scan A and B; around 5 kg, which the IDR was also adapted for in this
case (4).

Example II: Parenchymal studies

Figure 4 depicts a case where a patient is scanned twice with approximately

one year in between. The scans are both performed on a 3rd-generation
dual source CT scanner (Somatom Force, Siemens Healthineers, Forchheim,
Germany), the first at 120 kV and the second at 90 kV. This resulted in similar
attenuation values, because the CM dosing factor is adapted accordingly. A
30 kV tube voltage reduction resulted in a 30 % decrease of the dosing factor.
Injection time is held constant at 30 seconds. Therefore, the flow rate changes
with a changing body weight (6).

122

Vol_BM_productie.indd 122 02/02/2022 15:23:11

 Contrast Media Protocols: The 10-to-10 Rule

Figure 4. Two abdominal CT scans of the same patient in portal venous phase, per-
formed on the same scanner at different kV settings. The attenuation remains the same,
by adapting the total amount of CM administered to the kV setting and the patients’
body weight. The scan on the left is performed at 120 kV and the one on the right at
90 kV with one-year time difference on a 3rd-generation dual source CT scanner (6).

7
Conclusion

The 10-to-10 rule is an easy to use rule of thumb to adjust CM injection

protocols to varying tube voltages. When used in conjunction with patient
tailored injection protocols, this rule will aid in keeping image quality constant
and homogeneous throughout a patient population.

123

Vol_BM_productie.indd 123 02/02/2022 15:23:11

 Chapter 7

References

1. Goodman LR. The Beatles, the Nobel Prize, and CT scanning of the chest. Radiol
Clin North Am. 2010;48(1):1-7.
2. Lell MM, Wildberger JE, Alkadhi H, et al. Evolution in computed tomography: the
battle for speed and dose. Invest Radiol. 2015;50(9):629-44.
3. Lell MM, Kachelrieß M. Recent and Upcoming Technological Developments in
Computed Tomography: High Speed, Low Dose, Deep Learning, Multienergy.
Investigative Radiology. 9000;Publish Ahead of Print.
4. Hendriks BMF, Eijsvoogel NG, Kok M, et al. Optimizing pulmonary embolism
computed tomography in the age of individualized medicine: a prospective clinical
study. Invest Radiol. 2018;53(5):306-12.
5. Eijsvoogel NG, Hendriks BMF, Willigers JL, et al. Personalization of injection protocols
to the individual patient’s blood volume and automated tube voltage selection
(ATVS) in coronary CTA. PLoS One. 2018;13(9):e0203682.
6. Martens B, Wildberger JE, Hendriks BM, et al. A solution for homogeneous liver
enhancement in computed tomography: results from the COMpLEx trial. Under
Submission.
7. Mayer C, Meyer M, Fink C, et al. Potential for radiation dose savings in abdominal
and chest CT using automatic tube voltage selection in combination with automatic
tube current modulation. AJR Am J Roentgenol. 2014;203(2):292-9.
8. Papadakis AE, Damilakis J. Automatic Tube Current Modulation and Tube Voltage
Selection in Pediatric Computed Tomography: A Phantom Study on Radiation Dose
and Image Quality. Invest Radiol. 2019;54(5):265-72.
9. Kalra MK, Sodickson AD, Mayo-Smith WW. CT radiation: key concepts for gentle and
wise use. Radiographics. 2015;35(6):1706-21.
10. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
11. Kok M, Mihl C, Seehofnerova A, et al. Automated tube voltage selection for radiation
dose reduction in CT angiography using different contrast media concentrations
and a constant iodine delivery rate. AJR Am J Roentgenol. 2015;205(6):1332-8.
12. Kok M, Mihl C, Hendriks BM, et al. Patient comfort during contrast media
injection in coronary computed tomographic angiography using varying contrast
media concentrations and flow rates: results from the EICAR trial. Invest Radiol.
2016;51(12):810-5.
13. Hendriks BM, Kok M, Mihl C, et al. Individually tailored contrast enhancement in CT
pulmonary angiography. Br J Radiol. 2016;89(1061):20150850.
14. Mihl C, Wildberger JE, Jurencak T, et al. Intravascular enhancement with identical
iodine delivery rate using different iodine contrast media in a circulation phantom.
Invest Radiol. 2013;48(11):813-8.

124

Vol_BM_productie.indd 124 02/02/2022 15:23:11

 Contrast Media Protocols: The 10-to-10 Rule

15. Bae KT, Tao C, Gurel S, et al. Effect of patient weight and scanning duration on
contrast enhancement during pulmonary multidetector CT angiography. Radiology.
2007;242(2):582-9.
16. Mihl C, Kok M, Altintas S, et al. Evaluation of individually body weight adapted
contrast media injection in coronary CT-angiography. Eur J Radiol. 2016;85(4):830-6.
17. Martens B, Hendriks BMF, Eijsvoogel NG, et al. Individually body weight-adapted
contrast media application in computed tomography imaging of the liver at 90 kVp.
Invest Radiol. 2019;54(3):177-82.
18. Kondo H, Kanematsu M, Goshima S, et al. Body size indices to determine iodine
mass with contrast-enhanced multi-detector computed tomography of the upper
abdomen: does body surface area outperform total body weight or lean body
weight? Eur Radiol. 2013;23(7):1855-61.
19. Caruso D, De Santis D, Rivosecchi F, et al. Lean body weight-tailored iodinated
contrast injection in obese patient: boer versus james formula. Biomed Res Int.
2018;2018:8521893.
20. Hou SH, Bushinsky DA, Wish JB, et al. Hospital-acquired renal insufficiency: a
prospective study. Am J Med. 1983;74(2):243-8.
21. Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin
Pract. 2012;120(4):c179-84.
22. McDonald RJ, McDonald JS, Bida JP, et al. Intravenous contrast material-induced
nephropathy: causal or coincident phenomenon? Radiology. 2013;267(1):106-18.
23. Nijssen EC, Rennenberg RJ, Nelemans PJ, et al. Prophylactic hydration to protect 7
renal function from intravascular iodinated contrast material in patients at high risk
of contrast-induced nephropathy (AMACING): a prospective, randomised, phase 3,
controlled, open-label, non-inferiority trial. Lancet. 2017;389(10076):1312-22.
24. Kang SK, Huang WC, Pandharipande PV, Chandarana H. Solid renal masses: what
the numbers tell us. AJR Am J Roentgenol. 2014;202(6):1196-206.
25. Dyer R, DiSantis DJ, McClennan BL. Simplified imaging approach for evaluation of
the solid renal mass in adults. Radiology. 2008;247(2):331-43.
26. Kondo H, Kanematsu M, Goshima S, et al. Body size indexes for optimizing iodine
dose for aortic and hepatic enhancement at multidetector CT: comparison of total
body weight, lean body weight, and blood volume. Radiology. 2010;254(1):163-9.
27. Bae KT, Shah AJ, Shang SS, et al. Aortic and hepatic contrast enhancement with
abdominal 64-MDCT in pediatric patients: effect of body weight and iodine dose.
AJR Am J Roentgenol. 2008;191(5):1589-94.
28. Kok M, Mihl C, Hendriks BM, et al. Optimizing contrast media application in coronary
CT angiography at lower tube voltage: evaluation in a circulation phantom and sixty
patients. Eur J Radiol. 2016;85(6):1068-74.
29. Lell MM, Jost G, Korporaal JG, et al. Optimizing contrast media injection protocols in
state-of-the art computed tomographic angiography. Invest Radiol. 2015;50(3):161-7.

125

Vol_BM_productie.indd 125 02/02/2022 15:23:11

 Chapter 7

30. Canstein C, Korporaal JG. Reduction of contrast agent dose at low kV settings. In:
Siemens, ed. Forchheim, Germany; 2015.
31. Heiken JP, Brink JA, McClennan BL, et al. Dynamic incremental CT: effect of volume
and concentration of contrast material and patient weight on hepatic enhancement.
Radiology. 1995;195(2):353-7.
32. Brink JA, Heiken JP, Forman HP, et al. Hepatic spiral CT: reduction of dose of
intravenous contrast material. Radiology. 1995;197(1):83-8.
33. Canstein C, Korporaal JG. Reduction of contrast agent dose at low kV settings.
Siemens Healthineers White Paper. 2015.
34. Lell MM, Fleischmann U, Pietsch H, et al. Relationship between low tube voltage (70
kV) and the iodine delivery rate (IDR) in CT angiography: An experimental in-vivo
study. PLoS One. 2017;12(3):e0173592.

126

Vol_BM_productie.indd 126 02/02/2022 15:23:11

 Contrast Media Protocols: The 10-to-10 Rule

127

Vol_BM_productie.indd 127 02/02/2022 15:23:11

Vol_BM_productie.indd 128 02/02/2022 15:23:12
 CHAPTER
8
Book chapter:
Artificial intelligence based contrast
medium optimization

B. Martens, B.M.F. Hendriks, J.E. Wildberger, C. Mihl

Book title: Artificial Intelligence in Cardiothoracic Imaging

ISBN: 978-3-030-92086-9 Springer
1st edition 2022

Vol_BM_productie.indd 129 02/02/2022 15:23:13

 Chapter 8

Abstract

Contrast media (CM) application is important for the evaluation of the heart
with both computed tomography (CT) and magnetic resonance imaging (MRI).
In many hospitals around the world, CM is still used in a “one size fits all”
fashion, usually with a “safety margin” regarding CM volume, guaranteeing
sufficient enhancement even in the heavier patient. The primary reason for
using standard protocols instead of optimising CM for individual patients is
the fact that CM administration is still largely a manual action, time consuming
and regarded as sensitive for errors. Artificial intelligence (AI) techniques could
play a role in that respect. If AI can select the optimal CM injection protocol
for the specific patient, on a particular CT or MRI scanner for the clinical scan
indication, AI would improve both patient care and workflow. Different aspects
might extend the study or make the study more difficult, e.g. patient anxiety,
difficult venous access and/or an irregular heartbeat. In case these factors
could be taken into account when scheduling the examination, that would
further improve workflow. In addition, AI might help in further reducing CM,
scan time and – in case of CT – radiation dose. The position of AI with regard
to CM optimisation is not thoroughly studied yet, this chapter aims to offer
insights into possible future directions.

Keywords

Computed Tomography; Diagnostic Imaging; Coronary artery disease; Contrast

Media; Weight; Artificial Intelligence; Heart

130

Vol_BM_productie.indd 130 02/02/2022 15:23:13

 Artificial intelligence based contrast medium optimization

Background in scan and contrast media protocols

Coronary CT angiography (CCTA) has a very high negative predictive value

for ruling out coronary artery disease (99 %) (1). Sufficient intravascular
enhancement of the coronaries (> 325 Hounsfield Units [HU]) is necessary to
reach diagnostic accuracy (2). The k-edge of iodine lies at 33 keV and when the
tube voltage approaches the k-edge, the attenuation of iodine increases (3).
This makes it possible to increase the vascular or parenchymal enhancement by
decreasing the tube voltage of the scan acquisition. Different parameters can
be adapted to reach sufficient enhancement of the coronaries in each patient
on every occasion. First of all, the iodine delivery rate (IDR) is considered the
most decisive parameter for attenuation of vascular structures (4, 5). Second,
a body weight adapted CM injection protocol results in more homogeneous
enhancement of the coronary arteries, compared to a one size fits all protocol
(6). Lastly, the 10-to-10 rule optimizes the CM application even further based
on the tube voltage used: a 10 kV reduction should be followed by a 10 %
reduction in IDR and vice versa (7). To optimize CM timing, a test bolus or
bolus tracking technique can be used (8). For the test bolus method, multiple
low dose axial image slices are acquired at one point on the z-axis after the
injection of a small amount of CM. The test bolus is injected with the same flow
rate and concentration as the main bolus. This creates a time-enhancement 8
curve where the peak-enhancement can be determined, to find the optimal
timing for the scan. The time to peak is used as an estimated scan delay for
the scan with the main CM dose (9). For bolus tracking, CM is injected and
attenuation is measured in a predefined vessel until a threshold is reached.
Once the attenuation reaches the threshold, the scan starts after a predefined
fixed delay (10).

There are three different cardiac scan acquisition modes, ECG triggered or flash
(high-pitch mode), prospective ECG gated (step and shoot) and retrospective
ECG gating (helical). Based on heart rate and heart rhythm, a specific protocol
is chosen (11).

Iterative Reconstruction (IR) techniques, automated tube current modulation

(ATCM) and automated tube voltage selection (ATVS) are at present incorporated
in daily clinical routine. These techniques are performed semi-automatically,

131

Vol_BM_productie.indd 131 02/02/2022 15:23:13

 Chapter 8

as they have to be approved by a radiographer, and generally do not fall under

the topic of AI. CM injectors that combine scanner protocol information (e.g.
tube voltage and tube current selected by ATCM and ATVS) with the patient
characteristics and clinical question are a promising tool for automatically
optimising CM application but are yet to be developed and synchronised. To
date, individualising CM injection protocols is still largely a manual and time-
consuming action, which explains the preference for a standard “one size fits
all” protocol in many centres.

Artificial Intelligence for Contrast media optimisation

Because of technical developments the cardiac CT scan duration nowadays is

very short, ranging from several seconds to sub-second imaging. For imaging of
the coronary arteries, a fast scan acquisition is beneficial, as this makes imaging
of the whole heart possible within one heartbeat, subsequently reducing
radiation and CM dose. Faster scanners and shortened CM protocols result
in more critical scan timing, as the chance of missing the peak enhancement
increases, especially with respect to the vascular structures (9, 10, 12). AI could
improve coordination of scan and CM timing, resulting in optimized CM and
radiation dose in each individual patient.

The possibility to perform thoracic scans in a single heartbeat reduces motion

artefacts and cardiac structures can nowadays often be assessed on a regular
CT of the chest (12). Image quality might be sufficient to evaluate the heart,
although on a regular chest CT there might be not enough CM distribution in
the cardiac chambers. Suboptimal cardiac enhancement is not surprising, as
a dedicated cardiac scan is performed later after start of CM injection than a
regular CT of the chest (13). As a result, information regarding cardiopathies
and valvular changes stay hidden.

A triple rule out protocol has been advocated as a solution for better cardiac
image quality on chest CT’s (14). AI however may bypass the additional CM
volume of a triple rule out CM protocol and it could bypass the need for
CM entirely. A deep convolutional neural network (DCNN) has shown to be
beneficial in these areas too (15). Santini et al. developed a DCNN capable of
producing contrast-enhanced CT (CECT) images out of a non-contrast image

132

Vol_BM_productie.indd 132 02/02/2022 15:23:13

 Artificial intelligence based contrast medium optimization

of the thorax by mimicking the human visual system with the created network
(15). The left cardiac ventricle was delineated by the software on the non-
CECT images, leading to the synthesized CECT image. Non-contrast images
where compared to real CECT images and the synthesized CECT images. A
Normalized Mutual Information index (NMI) was used to quantify the capability
of the model to create a synthesized CECT close to the real CECT. The Dice
index is a statistical tool that compares the similarity between two datasets and
was used to evaluate the estimation of the cardiac ventricles. The synthesized
compared to the real CECT images gained a good similarity (NMI of 0.93 ±
0.03) and extraction of the left ventricle was possible (Dice = 0.88). Therefore
it was possible to subtract the left cardiac ventricle from a non-CECT image
and, in addition, to provide volumetric information of the heart on a regular
thoracic CT (15). Mannil et al. used texture analysis and machine learning to
detect a myocardial infarction on an unenhanced low radiation dose cardiac
CT (16). Both DCNN and machine learning turn out to be capable of retrieving
latent information from a non-CECT image. As information about the cardiac
ventricles can be subtracted from non-CECT images, it seems to be a small
step to generate an optimal contrast-enhanced coronary image of a CT with
reduced CM dosing or scans with alternative CM timing (such as chest CTs).

In current clinical practice, total CM volumes containing as little as 9 grams of 8

iodine can be used for CCTA with diagnostic image quality. Further reduction of
CM might not be necessary for the protection of kidney function (17). However,
it should be pointed out that reducing the amount of CM also decreases health
care costs. Furthermore, iodinated CM reaches the groundwater and has an
influence on the environment as well (18). AI could play a big role in aiding
wide acceptance and implementation of individualised CM protocols, by fully
automating the process of CM individualisation. This individualisation might be
beneficial for advanced data-characterisation algorithms, such as radiomics.
Differences in, for example, cardiac output and CM injection protocols influence
the distribution of the CM in tissue. In addition, differences in scan acquisition
parameters might influence spatial analysis (19). Therefore, differences in scan
and CM protocols can influence features derived from radiomics algorithms.
In order to optimally apply radiomics in the future, it seems vital that scan and
injection protocols become comparable between hospital and countries.

133

Vol_BM_productie.indd 133 02/02/2022 15:23:14

 Chapter 8

Takumi et al. used bolus tracking data of previous CT scans of the same patient
in dynamic liver imaging. The authors showed that a scan delay based on
previous bolus tracking data resulted in a similar CM enhancement. Previous
bolus tracking data could be used in following scans, shortening exam time
and reducing radiation exposure (20). However, this will be more challenging to
apply in cardiac imaging, as various different factors (e.g. medication, disease,
patient stress level and time of the day) will influence cardiac output.

Uniformity of the CM protocol throughout the different hospitals in the world,

could lead to optimal (coronary) enhancement in each patient, every time. In
addition, by individualising and optimising CM injection protocols all around the
world, attenuation of the targeted structures should become more uniform,
which could be beneficial for future AI techniques.

Artificial Intelligence for Contrast Media Injection Protocol selection

Scan duration and therefore the CM injection protocol for both the prospective
and retrospective ECG gated protocol are longer than for the ECG triggered
(flash) protocol to ensure optimal enhancement of the coronaries. Having three
different cardiac scan protocols makes the individualisation of scan and CM
protocols even more complex.

An irregular heartbeat can complicate the selection of the right protocol.

Problems in synchronisation of ECG signal with data acquisition sometimes
result in non-diagnostic image quality necessitating scan repetition, with no
guarantee of the second scan being of diagnostic image quality. AI could learn
to recognise specific heartbeat patterns and select the appropriate scan
protocol. Moreover, while determining the best scan protocol for the individual
patient, the AI model should be capable of selecting the appropriate CM
injection protocol on the available scanner for the specific question at hand.
Notably, bolus tracking with a fixed trigger delay could miss peak enhancement
in patients with a very high or very low cardiac output, because it does not take
patient specific cardiovascular parameters output into account (9, 10). AI and
machine learning algorithms could be applied to individualize trigger scan delay
when using bolus tracking. Hinzpeter et al. proposed to use at least four bolus
tracking enhancement values to evaluate the CM dynamics in the individual

134

Vol_BM_productie.indd 134 02/02/2022 15:23:14

 Artificial intelligence based contrast medium optimization

patient. This information can be used to match the data to the available
enhancement curves in an online database containing different arterial blood
circulation curves. Thereby the optimal scan timing could be determined based
on the best suited enhancement curve for each patient (10, 21).

Ideally, AI will suggest the most appropriate protocol so suboptimal image

quality and repeated scanning can be a thing of the past. In addition, it would be
of great value if it were possible to predict whether a prospectively ECG gated
acquisition will be executed smoothly. For example, an AI algorithm capable of
predicting that an extra systole will lead to a longer scan time, can adapt the
CM bolus mid-scan to ensure sufficient enhancement throughout the whole
scan duration.

AI-based Image reconstruction techniques

As mentioned before, decreasing the tube voltage is advantageous for both

radiation and CM dose. However, it comes at the expense of increased image
noise. Noise reduction algorithms offer dedicated opportunities in this respect.
A knowledge-based IR algorithm minimizes the difference between measured
raw CT data and the estimated image, by a penalty-based cost function (22).
It decreases image noise and increases the contrast-to-noise ratio (CNR) (23). 8
Wang et al. evaluated an AI-based noise reduction algorithm based on a DCNN
(24). Their AI-based noise reduction algorithm based on 40 CT scans of the
aorta at normal dose and at a simulated low dose. The dedicated algorithm
learned that the output was input minus noise and, in this way, was able to
generate noise reduced images. The algorithm improved image quality of the
aorta, facilitating a 50 % CM dose reduction (24).

Another possibility for radiation and CM reduction lies in the large portion of
follow-up CT scans performed on a daily basis, for example for oncological
patients. These patients undergo repeated scanning of the same anatomical
region. Therefore, shared anatomical information between those scans is
available. This information can be used in an IR algorithm to significantly improve
the diagnostic image quality (25). In the future, it might become possible to use
the anatomical information from scans performed for a different indication, for
a scheduled cardiac scan of the same patient, thereby reducing both radiation

135

Vol_BM_productie.indd 135 02/02/2022 15:23:14

 Chapter 8

and CM dose. This could lead to both scan and CM protocol optimisation as
well as improvements in workflow.

Non-calcified and low attenuating plaques are more prone to future events than
calcified plaques, as calcifications might stabilise the coronary plaque (26). In
addition, positive remodelling, spotty calcifications and the napkin-ring sign are
indicators of a high-risk plaque (27). Occasionally, some rather big differences
in interpretation between readers can be present and especially blooming
artefacts hamper the interpretation of soft plaques. In that respect, subtraction
techniques might be promising for reducing these blooming artefacts (28).
Differences in scan quality, CM application but also the experience of the reader
and the subtlety of the findings will influence subjective evaluation of the CCTA
(29). AI techniques could be helpful in standardising plaque interpretation and
might even go a step further in determining plaque characteristics (27, 30).

Workflow

Scheduling a CT or MRI scan is often a job done by the supportive staff in the
radiology department. It is time consuming and prone to errors. AI algorithms
could be used to plan and schedule this medical care (31). The selection of the
correct imaging modality, scan and CM injection protocol, as well as the ideal
study date (e.g. should the scan be performed within an hour or a month) are
parameters to keep in mind, when scheduling an exam. A task that might be
difficult and prone to changes, but could be rather easy for an AI algorithm.
This automated scheduling could involve prioritising patients and selecting the
optimal imaging technique (e.g. ultrasound, CT or MRI) (31).

Information from previous studies could be taken into account. Anxious

patients, known risk factors for CM related unfavourable outcomes, difficulty
with obtaining venous access or a known irregular heartbeat (with exception
of paradoxical atrial fibrillation, which is often not predictable) could be known
upfront. This information could be used to select specific time slots for the
requested CT or MRI exam. Predicting the scan type and scan duration correctly
will improve efficient use of scanners and staff (21). Altogether, this optimized
workflow will lead to improved patient care.

136

Vol_BM_productie.indd 136 02/02/2022 15:23:14

 Artificial intelligence based contrast medium optimization

Red flags

The minimal diagnostic vascular enhancement for the coronary arteries,

aorta and the pulmonary arteries are well studied (32). An attenuation below
the threshold might result in non-diagnostic image quality, but when the
enhancement is too high, CM can lead to artifacts mimicking relevant findings,
making image quality suboptimal as well. For each target (vascular structure), it
is possible to define an attenuation range for each CT exam. In case the limit is
not reached or if the diagnostic HU level is exceeded, the study is regarded as
suboptimal. It might be possible to exclude gross pathologies on images with
suboptimal attenuation of the designated structures, but there is an increased
risk that significant findings might be missed or misinterpreted. Factors such as
motion artefacts, irregular heartbeat and unintentional incorrect scanner and/
or injector settings might result in an unanticipated insufficient image quality
(33). Ideally, an algorithm could approve image quality before the patient leaves
the scanner room. The program should be capable of determining whether the
coronary attenuation is sufficient and if artefacts are present that influence the
interpretation significantly.

On some occasions an outpatient CT or MRI scan might be in a worklist for

several days, while the scan shows unexpected pathology requiring immediate 8
attention. AI could be trained to recognise and ‘flag’ a scan when unexpected
clinically significant findings are indicated that should be reviewed right away.
For example, a significant unstable coronary plaque, an aneurysm or dissection
of the ascending aorta or an imminent cardiac tamponade could trigger an
alarm signal notifying the radiologist to look at this exam immediately.

Summary

AI might be able to contribute to quality insurance by determining an

individualized CM, scan and reconstruction technique in each patient. The
scan could automatically be followed by an attenuation check of the targeted
structure and a warning signal in case of any pathology present that is in need
of immediate attention (figure 1). One step further, it would be valuable if AI
could determine which parameters to adjust and to what extent to reach a

137

Vol_BM_productie.indd 137 02/02/2022 15:23:14

 Chapter 8

sufficient image quality when the first or any previous scan was regarded as
non-diagnostic.

Figure 1. The role of artificial intelligence (AI) in contrast optimisation. The process
starts with scan indication, followed by heart rate, which will influence which scan
protocol is most appropriate. Next, the corresponding contrast media (CM) protocol
is selected and aligned with individual patient parameters. The process ends with a
quality check. If quality is insufficient, the process will iterate to improve scan and CM
protocols based on errors and specific characteristics from previous exams. After image
acquisition the process is followed by an automated search for urgent findings which
may require immediate attention.

AI could be useful in individualising CM protocols based on patient and scanner

characteristics as well as on suspected clinical pathology to improve both,
daily clinical care and workflow. Although, in that respect, the question rises
if this might be applicable in the near future because of legal considerations:
who is taking the responsibilities if something goes wrong? This leads to an
ethical discussion that will have to take place before AI can become widely
incorporated in medical imaging. The example of self-driving cars sets out
the grey area clearly. Who is responsible when a bystander gets hurt, can the
self-driving car be responsible? Compared to a human, the car will probably
not be able to make a morally guided decision, when an accident is about to
happen. Because of these discussions, the introduction of self-driving cars
did not happen yet (34). AI taking over the radiologists’ job is ethically not so
easy either.

138

Vol_BM_productie.indd 138 02/02/2022 15:23:14

 Artificial intelligence based contrast medium optimization

One frequently reappearing topic of discussion involves AI taking over the

radiologists’ job. According to Geoffrey Hinton, no more radiologists should be
trained, because deep learning would outperform the radiologist within 5 years,
stated in 2016 in Toronto (35). As long as AI is seen as a potential threat, we
might miss all the opportunities it has in store for us. This chapter has pointed
out the possibilities of a DCNN, machine learning and deep learning as a tool
for CM optimalisation in both CT and MRI. These kinds of applications could
be used to improve workflow, image quality and even diagnostic capabilities.
The future of radiology is bright, if we adopt the opportunities AI offers early.

139

Vol_BM_productie.indd 139 02/02/2022 15:23:14

 Chapter 8

References

1. Plank F, Friedrich G, Dichtl W, Klauser A, Jaschke W, Franz WM, et al. The diagnostic
and prognostic value of coronary CT angiography in asymptomatic high-risk
patients: a cohort study. Open Heart. 2014;1(1):e000096.
2. Eijsvoogel NG, Hendriks BMF, Nelemans P, Mihl C, Willigers J, Martens B, et al.
Personalization of CM Injection Protocols in Coronary Computed Tomographic
Angiography (People CT Trial). Contrast Media Mol Imaging. 2020;2020:5407936.
3. Fleischmann U, Pietsch H, Korporaal JG, Flohr TG, Uder M, Jost G, et al. Impact of
Contrast Media Concentration on Low-Kilovolt Computed Tomography Angiography:
A Systematic Preclinical Approach. Invest Radiol. 2018;53(5):264-70.
4. Kok M, Mihl C, Hendriks BM, Altintas S, Kietselaer BL, Wildberger JE, et al. Optimizing
contrast media application in coronary CT angiography at lower tube voltage:
Evaluation in a circulation phantom and sixty patients. Eur J Radiol. 2016;85(6):1068-
74.
5. Rengo M, Dharampal A, Lubbers M, Kock M, Wildberger JE, Das M, et al. Impact
of iodine concentration and iodine delivery rate on contrast enhancement in
coronary CT angiography: a randomized multicenter trial (CT-CON). Eur Radiol.
2019;29(11):6109-18.
6. Mihl C, Kok M, Altintas S, Kietselaer BL, Turek J, Wildberger JE, et al. Evaluation
of individually body weight adapted contrast media injection in coronary CT-
angiography. Eur J Radiol. 2016;85(4):830-6.
7. Martens B, Hendriks BMF, Mihl C, Wildberger JE. Tailoring contrast media protocols
to varying tube voltages in vascular and parenchymal CT imaging: The 10-to-10 rule.
Invest Radiol. 2020.
8. Cademartiri F, Nieman K, van der Lugt A, Raaijmakers RH, Mollet N, Pattynama
PM, et al. Intravenous contrast material administration at 16-detector row helical
CT coronary angiography: test bolus versus bolus-tracking technique. Radiology.
2004;233(3):817-23.
9. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
10. Hinzpeter R, Eberhard M, Gutjahr R, Reeve K, Pfammatter T, Lachat M, et al. CT
Angiography of the Aorta: Contrast Timing by Using a Fixed versus a Patient-specific
Trigger Delay. Radiology. 2019;291(2):531-8.
11. Machida H, Tanaka I, Fukui R, Shen Y, Ishikawa T, Tate E, et al. Current and Novel
Imaging Techniques in Coronary CT. Radiographics : a review publication of the
Radiological Society of North America, Inc. 2015;35(4):991-1010.
12. Lell MM, Wildberger JE, Alkadhi H, Damilakis J, Kachelriess M. Evolution in Computed
Tomography: The Battle for Speed and Dose. Invest Radiol. 2015;50(9):629-44.

140

Vol_BM_productie.indd 140 02/02/2022 15:23:14

 Artificial intelligence based contrast medium optimization

13. Frauenfelder T, Appenzeller P, Karlo C, Scheffel H, Desbiolles L, Stolzmann P, et al.

Triple rule-out CT in the emergency department: protocols and spectrum of imaging
findings. Eur Radiol. 2009;19(4):789-99.
14. Kidoh M, Nakaura T, Nakamura S, Namimoto T, Nozaki T, Sakaino N, et al. Contrast
material and radiation dose reduction strategy for triple-rule-out cardiac CT
angiography: feasibility study of non-ECG-gated low kVp scan of the whole chest
following coronary CT angiography. Acta Radiol. 2014;55(10):1186-96.
15. Santini G, Zumbo L, Martini N, Valvano G, Leo A, Ripoli A, et al. Synthetic contrast
enhancement in cardiac CT with Deep Learning 2018 [Available from: https://arxiv.
org/pdf/1807.01779.pdf.
16. Mannil M, von Spiczak J, Manka R, Alkadhi H. Texture Analysis and Machine
Learning for Detecting Myocardial Infarction in Noncontrast Low-Dose Computed
Tomography: Unveiling the Invisible. Invest Radiol. 2018;53(6):338-43.
17. Nijssen EC, Rennenberg RJ, Nelemans PJ, Essers BA, Janssen MM, Vermeeren
MA, et al. Prophylactic hydration to protect renal function from intravascular
iodinated contrast material in patients at high risk of contrast-induced nephropathy
(AMACING): a prospective, randomised, phase 3, controlled, open-label, non-
inferiority trial. Lancet. 2017;389(10076):1312-22.
18. Kormos JL, Schulz M, Ternes TA. Occurrence of iodinated X-ray contrast media and
their biotransformation products in the urban water cycle. Environ Sci Technol.
2011;45(20):8723-32.
19. Ibrahim A, Primakov S, Beuque M, Woodruff HC, Halilaj I, Wu G, et al. Radiomics for
precision medicine: Current challenges, future prospects, and the proposal of a new
framework. Methods. 2020. 8
20. Takumi K, Fukukura Y, Shindo T, Kumagae Y, Tateyama A, Kamiyama T, et al. Feasibility
of a fixed scan delay technique using a previous bolus tracking technique data for
dynamic hepatic CT. Eur J Radiol. 2012;81(11):2996-3001.
21. Eberhard M, Alkadhi H. Machine Learning and Deep Neural Networks: Applications in
Patient and Scan Preparation, Contrast Medium, and Radiation Dose Optimization.
J Thorac Imaging. 2020;35 Suppl 1:S17-S20.
22. Yuki H, Utsunomiya D, Funama Y, Tokuyasu S, Namimoto T, Hirai T, et al. Value of
knowledge-based iterative model reconstruction in low-kV 256-slice coronary CT
angiography. J Cardiovasc Comput Tomogr. 2014;8(2):115-23.
23. Iyama Y, Nakaura T, Yokoyama K, Kidoh M, Harada K, Oda S, et al. Low-Contrast
and Low-Radiation Dose Protocol in Cardiac Computed Tomography: Usefulness of
Low Tube Voltage and Knowledge-Based Iterative Model Reconstruction Algorithm.
Journal of computer assisted tomography. 2016;40(6):941-7.
24. Wang Y, Yu M, Wang M, Wang Y, Kong L, Yi Y, et al. Application of Artificial Intelligence-
based Image Optimization for Computed Tomography Angiography of the Aorta
With Low Tube Voltage and Reduced Contrast Medium Volume. J Thorac Imaging.
2019;34(6):393-9.

141

Vol_BM_productie.indd 141 02/02/2022 15:23:14

 Chapter 8

25. Willemink MJ, Noel PB. The evolution of image reconstruction for CT-from filtered
back projection to artificial intelligence. Eur Radiol. 2019;29(5):2185-95.
26. von Knebel Doeberitz PL, De Cecco CN, Schoepf UJ, Albrecht MH, van Assen M, De
Santis D, et al. Impact of Coronary Computerized Tomography Angiography-Derived
Plaque Quantification and Machine-Learning Computerized Tomography Fractional
Flow Reserve on Adverse Cardiac Outcome. Am J Cardiol. 2019;124(9):1340-8.
27. Kolossvary M, De Cecco CN, Feuchtner G, Maurovich-Horvat P. Advanced
atherosclerosis imaging by CT: Radiomics, machine learning and deep learning. J
Cardiovasc Comput Tomogr. 2019;13(5):274-80.
28. Duarte Conde MP, de Korte AM, Meijer FJA, Aquarius R, Boogaarts HD, Bartels R,
et al. Subtraction CTA: An Alternative Imaging Option for the Follow-Up of Flow-
Diverter-Treated Aneurysms? AJNR Am J Neuroradiol. 2018;39(11):2051-6.
29. Saur SC, Alkadhi H, Stolzmann P, Baumuller S, Leschka S, Scheffel H, et al. Effect
of reader experience on variability, evaluation time and accuracy of coronary
plaque detection with computed tomography coronary angiography. Eur Radiol.
2010;20(7):1599-606.
30. van Assen M, Varga-Szemes A, Schoepf UJ, Duguay TM, Hudson HT, Egorova S, et al.
Automated plaque analysis for the prognostication of major adverse cardiac events.
Eur J Radiol. 2019;116:76-83.
31. Spyropoulos CD. AI planning and scheduling in the medical hospital environment.
Artif Intell Med. 2000;20(2):101-11.
32. Scholtz JE, Ghoshhajra B. Advances in cardiac CT contrast injection and acquisition
protocols. Cardiovasc Diagn Ther. 2017;7(5):439-51.
33. Ghekiere O, Salgado R, Buls N, Leiner T, Mancini I, Vanhoenacker P, et al. Image
quality in coronary CT angiography: challenges and technical solutions. Br J Radiol.
2017;90(1072):20160567.
34. Lau A. The Ethics of Self-Driving Cars. Should cars determine if you live or die?
2020 [Available from: https://towardsdatascience.com/the-ethics-of-self-driving-
cars-efaaaaf9e320
35. Mukherjee S. A.I Versus M.D. 2017 [Available from: http://www.newyorker.com/
magazine/2017/04/03/ai-versus-md.

142

Vol_BM_productie.indd 142 02/02/2022 15:23:14

 Artificial intelligence based contrast medium optimization

143

Vol_BM_productie.indd 143 02/02/2022 15:23:14

Vol_BM_productie.indd 144 02/02/2022 15:23:16
 CHAPTER
9
General discussion

Vol_BM_productie.indd 145 02/02/2022 15:23:17

 Chapter 9

The aim of this thesis was to optimize and individualize radiation dose and
contrast media (CM) injection protocols in abdominal computed tomography
(CT). Image quality is influenced by scanner-related (e.g. tube voltage, tube
current, slice reconstruction, scan delay and kernel), CM-related (e.g. CM
volume, flow rate, concentration, iodine delivery rate (IDR), viscosity, needle
gauge and saline chaser) and patient-related (e.g. weight, body mass index
(BMI), blood volume, heart rate, cardiac output and breath hold) parameters
(1-15). All of these should be taken into account when performing a CT scan,
as they not only affect image quality but also each other. Various parameters
were analysed separately to gain better insight into the different aspects of
scanner related, CM related and patient related factors, with the ultimate goal
being an individualized optimized protocol integrating all pertinent parameters.

In chapter 2, it was shown that an individualized CM injection protocol,

adapting both CM volume and flow rate to patient body weight, results in more
homogeneous enhancement of liver parenchyma than a standard CM injection
protocol (figure 1). Furthermore, a body weight adapted CM injection protocol
results in CM volume reduction in a large percentage of the population. This CM
reduction was comparable to those achieved in similar studies in the setting of
cardiac and pulmonary embolism (9, 12).

Even though total body weight is an easy-to-use approach in clinical practice,

previous studies found that lean body weight (LBW) may be more precise. This
is because fat contains less blood vessels than other tissue and does not play
an important role in CM distribution (16-18).

Furthermore, adapting CM volume to patient body weight is not sufficient for

reaching optimal image quality. Tube voltage must be taken into account as well,
because as it approaches the 33 keV k-edge of iodine decreasing tube voltage
results in increased CM attenuation.

The COMpLEx randomised controlled trial detailed in Chapter 3 puts this

theory into practice. Participants were randomly assigned into four groups
according to body weight-adapted CM dosing factor and tube voltage. Group
1 received 0.521 g I/kg CM at a tube voltage of 120 kV (based on the landmark
paper by Heiken et al. (2)).

146

Vol_BM_productie.indd 146 02/02/2022 15:23:17

 General discussion

Figure 1. When contrast (CM) volume and flow rate are adapted to patient body weight,
weight-dependent attenuation is eliminated. Homogeneous enhancement of the liver
parenchyma is achieved without compromising image quality.

Group 2 received the same CM dosage at a lower tube voltage of 90 kV. Groups
3 and 4 received CM dosages adapted according to the 10-to-10 rule, which pairs
a 10 kV reduction in tube voltage with a 10 % decrease in CM dose (based on
Canstein and Korporaal (19)). Group 3 was scanned at 100 kV (20 kV reduction
compared to Group 1), with 0.417 g I/kg CM (20 % CM reduction), and group 4 9
was scanned at 90 kV (-30 kV), with 0.365 g I/kg CM (-30 %). The results showed
that the proposed 10-to-10 rule is an easily reproducible method for achieving
homogeneous enhancement in portal venous CT of the liver throughout the
patient population, irrespective of patient body weight or tube voltage. This is
illustrated in Figure 2, which shows images of two patients undergoing regular
CT scans for oncological follow-up: even though parameters differ between the
first and second scans - both tube voltage and contrast volume are reduced by
30% - comparable attenuation of the liver parenchyma is achieved.

Automated tube current modulation (ATCM) and automated tube voltage

selection (ATVS) techniques individualize radiation dose based on patient body
habitus and the clinical question, but CM injection protocols are still mainly
set manually. A connection between scanner and CM injector would enable

147

Vol_BM_productie.indd 147 02/02/2022 15:23:17

 Chapter 9

automation of the 10-to-10 rule, making it easier to adapt protocols to both

tube voltage and patient, decreasing time spent by the technician in setting up
the CM protocol, and eventually decreasing costs (e.g. by being able to perform
more scans on a daily basis).

Figure 2. Two repeat scans of two oncological follow-up patients (1 and 2).

Repeat scans were done 6-12 months apart using different scan protocols. First scans (A &
B) were performed with standard tube voltage of 120 kV and a body weight adapted contrast
injection protocol. The second scans (C & D), were performed with a 30 kV tube voltage
reduction and a corresponding 30 % decrease in dosing factor, as per the 10-to-10 rule.
Enhancement of the liver parenchyma is comparable between scans of the same patient.

Since LBW may be a more reliable than total body weight for CM injection
protocol adaptation, it would be interesting to study the performance of the
10-to-10 rule when using LBW.

148

Vol_BM_productie.indd 148 02/02/2022 15:23:17

 General discussion

Image quality is not only influenced by CM injection but also radiation dose
protocols. The radiation dose can be decreased by reducing tube voltage
(in kV), tube current (in mAs) or both (19, 20). In chapter 4, reconstruction
software was used to show that a tube current reduction of 10-40 % was
possible with iterative reconstruction (IR) strength 4. The highest objective
and subjective image quality was achieved with a 10 % mAs reduction, but a
40 % reduction still maintained sufficient image quality. Results were based on
pairwise intra-patient comparisons, using raw CT image data to mimic lower
tube current scans (by increasing noise) at different IR strengths. In absence
of a reliable image quality standard, various objective and subjective image
quality parameters were used. For objective image quality, signal-to-noise ratio
(SNR) and contrast-to-noise ratio (CNR) are most often used in the literature
to illustrate image quality, but there is no consensus on threshold or cut-off
values (21-25). Subjective image quality, as the name implies, is subjective, and
while some readers prefer more noise, others may have a preference for a
smoother appearance of depicted organs (26). Thus, results on image quality
cannot be considered hard outcomes, and illustrate the need for an objective
parameter that can be used to reliably and consistently assess image quality
in a generalisable way.

Aside from CM dose protocols based on body weight, tube voltage, tube
current and IR strength, other patient-related parameters are of importance.
In Chapter 5, adaptation of either CM dose or radiation dose, depending 9
on patient’s age and renal function, is investigated. Because tube voltage is
dependent not only on the clinical question and a user set CNR but also on
patient body composition, a higher tube voltage is used in heavier patients and
a lower tube voltage in leaner patients to reach the same CNR. However, CNR is
also affected by the amount of CM used. Thus, a similar CNR could be achieved
by simultaneously decreasing radiation dose and increasing CM dose, or vice
versa. This can be achieved using image-task-dependent optimisation settings
(slider levels) (32): position 11 leads to an increase in noise (-26 % radiation dose)
and an increase in CM dose (+26 % volume), and in position 3 noise is low and
CNR is based solely on the fat-water contrast(27). This protocol can be used to
achieve a -16 % reduction in CM dose, at the cost of a 37 % increase in radiation
dose to generate the same CNR.

149

Vol_BM_productie.indd 149 02/02/2022 15:23:17

 Chapter 9

The study detailed in Chapter 5 evaluates the above in 6 Gottingen minipigs.

The pigs were scanned in early arterial, portal venous and delayed phase with
three different scan and CM injection protocols. The standard protocol was a
slider position 7 and CM dose of 0.350 gI/kg, CM dose saving was slider position
3 with 0.294 gI/kg, and position 11 was used for radiation dose saving, with an
increase in CM volume to 0.441 gI/kg. Results showed no significant differences
between groups regarding SNR, CNR and subjective image quality. Only noise
was rated significantly higher in the radiation dose saving group. These results
are promising for including age and kidney function in further individualisation
of scan and CM injection protocols.

One other factor must be considered for a complete picture of image quality
and optimalization of protocols: patient comfort. Not only is this important to
the patient in question, but an uncomfortable patient will also affect image
quality by physiological stress reactions and the inability to be still. Furthermore,
such a patient may be afraid to come back, further increasing interference every
repeat scan. Such factors may ruin the timing of both scan and CM injection
protocol, not to mention planned schedules. In Chapter 6 the effect of pre-
warming CM on patient comfort and pain is evaluated. CM was injected either
at room temperature (~23°C [~73°F]) or pre-warmed to body temperature (37°C
[99°F]). Results showed that iodinated CM at room temperature was not inferior
to pre-warmed CM in patient comfort in abdominal CT imaging. Furthermore,
pre-warming CM did not lead to an increase in image quality, safety and/or
patient comfort. The flow rate in the study was low, but in our hospital more
than 90 % of scans performed between 2013 and 2019 were done with a similar
flow rate (<6 ml/s). Regarding optimization and individualization of radiation and
CM protocols, pre-warming CM might therefore no longer be a pre-requisite in
standard state-of-the art abdominal injection protocols in daily clinical routine.

Chapter 7 combines the most important protocol individualization aspects in

both vascular and parenchymal studies, including the 10-to-10 rule, based on
various studies performed by our group over the last years (6, 7, 9-13, 15). IDR
is considered the most decisive factor of the CM injection protocol in vascular
studies, but in parenchymal studies total CM volume is leading (1). Therefore,
in vascular studies a tube voltage reduction of 10 kV should be followed by a

150

Vol_BM_productie.indd 150 02/02/2022 15:23:17

 General discussion

10 % decrease in IDR, whereas in abdominal imaging a 10 kV reduction should

be accompanied by a 10 % decrease in dosing factor (in g I/s).

Figure 3 provides an overview of how to reach individualized optimized scan

and CM injection protocols in abdominal imaging, integrating all pertinent
parameters evaluated in this thesis. The process starts with the patient. Ideally
the patients’ body weight is measured on a calibrated weighing scale available
in the scanner room. The measured body weight is used to individualize the
CM injection protocol. Subsequently, the slider position is chosen based on
the age of the patient and/or kidney function. ATCM and ATVS techniques are
used to individualize the radiation dose to the individual patient body habitus.
Based on the 10-to-10 rule the CM injection protocol is adapted to both body
weight and the ATVS selected tube voltage. Contrast material can be injected at
either room temperature or pre-warmed, without negatively influencing image
quality, safety, or patient comfort. By reconstructing images with IR strength
4, a tube current reduction of 10 – 40 % can be reached, resulting in excellent
to sufficient image quality. When taken separately, these different facets may
appear quite manageable and clear, but when put together they may become
confusing and difficult to handle. The combined theory may be considered
utopia for daily clinical practice if one expects technicians to take into account
all the different parameters and their interactions. Furthermore, too many
influencing parameters makes the individualized and optimized scan protocol
prone to error. AI might be necessary to make it workable. Chapter 8 inspects 9
how AI could determine individualized CM, scan and reconstruction parameters
for each patient. The focus lies on cardiac imaging, but suggestions are widely
applicable to other areas including abdominal imaging.

Future directions

The studies in the present thesis were all performed in the same hospital.
As a consequence CM types and concentrations as well as scanner vendors
between studies were similar. This limits generalizability of the results. Different
vendors use different techniques for optimizing image quality (4). The preferred
IR strength and the possible accompanying tube current reduction is vendor
specific.

151

Vol_BM_productie.indd 151 02/02/2022 15:23:17

 Figure 3. Overview of this thesis: custom-made computed tomography of the abdomen.

152

Vol_BM_productie.indd 152
Chapter 9

Customisation starts when a patient enters the scanner room. Body weight is ideally measured on a calibrated weighing scale to enable individualisation
of the contrast media (CM) injection protocol, after which the slider position is chosen, depending on patient age and/or kidney function (a tube current
reduction up to 40 % can be achieved by using iterative reconstruction [IR] strength 4). Automated tube current modulation (ATCM) and automated
tube voltage selection (ATVS) techniques enable the scanner to automatically determine optimal tube voltage based on patient characteristics. Finally,
CM volume is adapted to both patient body weight and the tube voltage used, bearing in mind the 10-to-10 rule. In this setting, iodinated CM can be
administered either pre-warmed or at room temperature. Artificial intelligence might be able to facilitate this customisation process in the future.

02/02/2022 15:23:17
 General discussion

However, chapter 4 provides information on how to find optimal IR strength

and tube current in each center, the method is again vendor dependent.
Not all vendors make the raw data available, which is required for the use of
reconstruction software. Furthermore, image-task-dependent optimisation
settings (slider levels) are only available on scanners of a specific vendor
type. By contrast the 10-to-10 rule is expected to be robust, as the physical
effect of the 33 keV edge of iodine is similar in all cases. Increasing CM iodine
concentration increases its viscosity, limiting generalisability of the results of
the CATCHY trial. It can only be stated that injecting CM with 300 mg iodine
per ml at room temperature or pre-warmed led to comparable image quality,
safety and patient comfort. These examples show the main limitations of the
present thesis. Preferably, the combined results need to be evaluated in a
large multi-center trial, taking into account the whole range of CM types and
concentrations as well as vendors.

Future research could focus on a variety of related themes. Most importantly,

a universal objective image quality parameter, able to capture all the different
pertinent aspects, should be developed. The want of such a parameter is a
common thread throughout the studies of this thesis. Collaboration between
clinical physicists and radiologists is essential here, and AI might be helpful in
objectifying aspects of image quality.

Another important aspect to study is which body size parameter is most reliable 9
for CM protocol individualisation. This has been done in the Chinese population,
and LBW appears to be most promising (16, 17, 28). Recently De Jong et al. used
an AI algorithm to calculate patient LBW and based CM protocols on weight
categories in Dutch patients: they found liver enhancement to be most strongly
associated with LBW compared to total body weight or BMI (18). Even though
these consistent results are encouraging, large prospective studies in more
diverse populations are necessary for firm conclusions to be drawn.

As present thesis shows, many of the aspects in individualizing radiation and

CM injection protocols have been investigated in CT imaging. However, in
magnetic resonance imaging (MRI) patients are still more likely to receive a
one-size-fits all protocol, or a protocol based on weight categories. This offers
a whole range of new research opportunities.

153

Vol_BM_productie.indd 153 02/02/2022 15:23:17

 Chapter 9

The CATCHY trial found little effect of CM temperature on patient comfort.

However, CM flow rate in the trial was relatively low (< 6 ml/s). It would be
interesting to evaluate the effect of CM temperature at higher flow rates.
Previous studies report increased rates of adverse events when certain types
of CM are not pre-warmed, and guidelines on safe use of contrast material
assume pre-warming CM will be helpful to minimize complications (7, 33-35,
37, 38). To the best of our knowledge, there is no logical explanation for such
an assumption. The largest study on the subject - a retrospective analysis of 24
830 power-injections - found no effect of pre-warming for iopamidol 300 but a
tripling of adverse events for 20°C iopamidol 370 (36). The authors concluded
that pre-warming may be beneficial in more viscous CM. However, the patient
comfort or mild adverse events were not included in the study, and room
temperatures may be higher than 20 degrees in clinical practice. During the
CATCHY trial room temperature CM was 23°C on average. Finally, both studies
kept the flow rates relatively low, < 6 ml/s. A large study comparing adverse
event rates and patient comfort at a wide range of flow rates and different CM
concentrations and with clinical practice reflecting temperatures may finally
provide a definite answer as to the utility of CM pre-warming.

AI is an emerging field in daily clinical practice and might be helpful in

individualizing both radiation and CM injection protocols. An AI driven
automatic attenuation check of the targeted structure could be next, followed
by automated pathology detection coupled with a warning signal in case of
acute pathology in need of immediate attention. Once a reliable, general
and consequent objective image quality parameter has been developed, AI
algorithms could be able to determine which parameters to adjust in cases of
insufficient image quality, avoiding non-diagnostic scans altogether.

Conclusion

The present thesis provides an overview of how different parameters can be

adjusted to individualize and optimize radiation and CM injection protocols. Not
only body weight and tube voltage are of the essence. Tube current, patient
comfort and age considerations may also contribute toward ideal image quality
for each patient, every time. Future research may focus on involving AI in this
process, to facilitate optimal integration of all the different facets and perhaps
even generate a result that is greater than the sum of its parts.

154

Vol_BM_productie.indd 154 02/02/2022 15:23:17

 General discussion

References

1. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
2. Heiken JP, Brink JA, McClennan BL, Sagel SS, Crowe TM, Gaines MV. Dynamic
incremental CT: effect of volume and concentration of contrast material and patient
weight on hepatic enhancement. Radiology. 1995;195(2):353-7.
3. Awai K, Hiraishi K, Hori S. Effect of contrast material injection duration and rate on
aortic peak time and peak enhancement at dynamic CT involving injection protocol
with dose tailored to patient weight. Radiology. 2004;230(1):142-50.
4. Lell MM, Wildberger JE, Alkadhi H, Damilakis J, Kachelriess M. Evolution in Computed
Tomography: The Battle for Speed and Dose. Invest Radiol. 2015;50(9):629-44.
5. Kok M, de Haan MW, Mihl C, Eijsvoogel NG, Hendriks BM, Sailer AM, et al.
Individualized CT Angiography Protocols for the Evaluation of the Aorta: A Feasibility
Study. Journal of vascular and interventional radiology : JVIR. 2016;27(4):531-8.
6. Kok M, Mihl C, Hendriks BM, Altintas S, Kietselaer BL, Wildberger JE, et al. Optimizing
contrast media application in coronary CT angiography at lower tube voltage:
evaluation in a circulation phantom and sixty patients. Eur J Radiol. 2016;85(6):1068-
74.
7. Kok M, Mihl C, Mingels AA, Kietselaer BL, Muhlenbruch G, Seehofnerova A, et al.
Influence of contrast media viscosity and temperature on injection pressure in
computed tomographic angiography: a phantom study. Invest Radiol. 2014;49(4):217-
23.
8. Kok M, Mihl C, Seehofnerova A, Turek J, Jost G, Pietsch H, et al. Automated tube
voltage selection for radiation dose reduction in CT angiography using different
contrast media concentrations and a constant iodine delivery rate. AJR Am J
9
Roentgenol. 2015;205(6):1332-8.
9. Mihl C, Kok M, Altintas S, Kietselaer BL, Turek J, Wildberger JE, et al. Evaluation
of individually body weight adapted contrast media injection in coronary CT-
angiography. Eur J Radiol. 2016;85(4):830-6.
10. Mihl C, Kok M, Wildberger JE, Altintas S, Labus D, Nijssen EC, et al. Coronary CT
angiography using low concentrated contrast media injected with high flow rates:
feasible in clinical practice. Eur J Radiol. 2015;84(11):2155-60.
11. Mihl C, Wildberger JE, Jurencak T, Yanniello MJ, Nijssen EC, Kalafut JF, et al.
Intravascular enhancement with identical iodine delivery rate using different iodine
contrast media in a circulation phantom. Invest Radiol. 2013;48(11):813-8.
12. Hendriks BM, Kok M, Mihl C, Bekkers SC, Wildberger JE, Das M. Individually
tailored contrast enhancement in CT pulmonary angiography. Br J Radiol.
2016;89(1061):20150850.

155

Vol_BM_productie.indd 155 02/02/2022 15:23:17

 Chapter 9

13. Hendriks BMF, Eijsvoogel NG, Kok M, Martens B, Wildberger JE, Das M. Optimizing
pulmonary embolism computed tomography in the age of individualized medicine:
a prospective clinical study. Invest Radiol. 2018;53(5):306-12.
14. Eijsvoogel NG, Hendriks BMF, Willigers JL, Martens B, Carati LF, Horehledova B,
et al. Personalization of injection protocols to the individual patient’s blood
volume and automated tube voltage selection (ATVS) in coronary CTA. PLoS One.
2018;13(9):e0203682.
15. Eijsvoogel NG, Hendriks BMF, Nelemans P, Mihl C, Willigers J, Martens B, et al.
Personalization of CM Injection Protocols in Coronary Computed Tomographic
Angiography (People CT Trial). Contrast Media Mol Imaging. 2020;2020:5407936.
16. Awai K, Kanematsu M, Kim T, Ichikawa T, Nakamura Y, Nakamoto A, et al. The optimal
body size index with which to determine iodine dose for hepatic dynamic CT: a
prospective multicenter study. Radiology. 2016;278(3):773-81.
17. Kondo H, Kanematsu M, Goshima S, Tomita Y, Kim MJ, Moriyama N, et al. Body
size indexes for optimizing iodine dose for aortic and hepatic enhancement at
multidetector CT: comparison of total body weight, lean body weight, and blood
volume. Radiology. 2010;254(1):163-9.
18. de Jong DJ, Veldhuis WB, Wessels FJ, de Vos B, Moeskops P, Kok M. Towards
Personalised Contrast Injection: Artificial-Intelligence-Derived Body Composition
and Liver Enhancement in Computed Tomography. J Pers Med. 2021;11(3).
19. Kalra MK, Maher MM, Toth TL, Schmidt B, Westerman BL, Morgan HT, et al.
Techniques and applications of automatic tube current modulation for CT. Radiology.
2004;233(3):649-57.
20. Martin CJ, Sookpeng S. Setting up computed tomography automatic tube current
modulation systems. J Radiol Prot. 2016;36(3):R74-r95.
21. Goshima S, Kanematsu M, Noda Y, Kondo H, Watanabe H, Kawada H, et al.
Determination of optimal intravenous contrast agent iodine dose for the detection
of liver metastasis at 80-kVp CT. Eur Radiol. 2014;24(8):1853-9.
22. Holmquist F, Soderberg M, Nyman U, Falt T, Siemund R, Geijer M. 80-kVp hepatic
CT to reduce contrast medium dose in azotemic patients: a feasibility study. Acta
Radiol. 2020;61(4):441-9.
23. Miyoshi K, Onoda H, Tanabe M, Nakao S, Higashi M, Iida E, et al. Image quality in dual-
source multiphasic dynamic computed tomography of the abdomen: evaluating the
effects of a low tube voltage (70 kVp) in combination with contrast dose reduction.
Abdom Radiol (NY). 2020;45(11):3755-62.
24. Akagi M, Nakamura Y, Higaki T, Narita K, Honda Y, Zhou J, et al. Deep learning
reconstruction improves image quality of abdominal ultra-high-resolution CT. Eur
Radiol. 2019;29(11):6163-71.
25. Choi SJ, Ahn SJ, Park SH, Park SH, Pak SY, Choi JW, et al. Dual-source abdominopelvic
computed tomography: comparison of image quality and radiation dose of 80 kVp
and 80/150 kVp with tin filter. PLoS One. 2020;15(9):e0231431.

156

Vol_BM_productie.indd 156 02/02/2022 15:23:17

 General discussion

26. Geyer LL, Schoepf UJ, Meinel FG, Nance JW, Jr., Bastarrika G, Leipsic JA, et al. State
of the art: iterative CT reconstruction techniques. Radiology. 2015;276(2):339-57.
27. Euler A, Taslimi T, Eberhard M, Kobe A, Reeve K, Zimmermann A, et al. Computed
Tomography Angiography of the Aorta-Optimization of Automatic Tube Voltage
Selection Settings to Reduce Radiation Dose or Contrast Medium in a Prospective
Randomized Trial. Invest Radiol. 2021;56(5):283-91.
28. Kondo H, Kanematsu M, Goshima S, Watanabe H, Onozuka M, Moriyama N, et al.
Aortic and hepatic enhancement at multidetector CT: evaluation of optimal iodine
dose determined by lean body weight. Eur J Radiol. 2011;80(3):e273-7.

157

Vol_BM_productie.indd 157 02/02/2022 15:23:18

Vol_BM_productie.indd 158 02/02/2022 15:23:19
 CHAPTER
10
Summary
Nederlandse samenvatting

Vol_BM_productie.indd 159 02/02/2022 15:23:20

 Chapter 10

Since the invention of the computed tomography (CT) scanner in 1971, contrast
media (CM) injection protocols, software, and scanners have rapidly evolved.
In the beginning, a one-size-fits all scan protocol was applied: administered
radiation and CM doses were similar for each patient. However, as scanners
evolved it became possible to use different tube current and tube voltage
settings based on individual body composition. Automated tube current
modulation (ATCM) and automated tube voltage selection (ATVS) techniques
optimize radiation dose based on patient characteristics as well as a user
set image quality. Previous studies showed that in vascular studies the CM
injection protocol is mainly determined by the iodine delivery rate (IDR), while
in parenchymal studies total CM volume is most decisive. A decrease in tube
voltage will reduce radiation dose, but due to the 33 keV-edge of iodine it will
also result in increased attenuation. Therefore, a reduction in tube voltage is
advantageous for both radiation and CM dose. The downside of decreasing
tube voltage is an increase in image noise, which is why radiologists must work
with the delicate balance reflected in the “as low as reasonably achievable”
(ALARA) principle. The aim of the present thesis was to investigate this balance
so as to provide guidance for individualisation of both radiation and CM dose,
based on the clinical question and patient characteristics, and to obtain optimal
image quality in each patient, every time.

In chapter 2 a body weight adapted CM injection protocol in abdominal imaging

is introduced. As has previously been shown in cardiac and pulmonary artery
imaging, an individualized protocol based on body weight results in a more
homogeneous enhancement of the target structure (the liver), compared to a
one-size-fits all protocol.

The hypothesis of chapter 3 was that a body weight-based CM injection

protocol adapted to the tube voltage used would result in homogeneous
enhancement of the liver parenchyma across patients. In this double-blinded
randomized controlled trial, 256 patients were randomly assigned to one of
the four groups. In group 1, the reference group, the presumed gold standard
scanning and CM protocols were used: 120 kV and 0.521 g I/kg. In group 2, the
tube voltage was reduced to 90 kV, but CM administration was maintained
as for group 1. In group 3, the tube voltage was reduced to 100 kV, i.e., 20 kV
less than group 1, and CM was reduced by a corresponding 20 % to 0.417 g I/

160

Vol_BM_productie.indd 160 02/02/2022 15:23:20

 Summary

kg. Group 4 received a 90 kV scan protocol (-30 kV), and a corresponding CM

reduction to 0.365 g I/kg (-30 %). The results confirmed the hypothesis: the
weight and tube voltage-based CM injection protocols used in groups 1, 3 and 4
led to homogeneous enhancement of the liver across patients in portal venous
phase abdominal imaging.

In chapter 4 the optimal iterative reconstruction (IR) strength and tube current
for abdominal imaging are investigated using reconstruction software. Pairwise
intra-patient comparisons showed that IR strength 4 led to the best subjective
image quality, while a 10 to 40 % reduction in tube current was possible without
compromising the objective and subjective image quality.

In the previous chapters, optimisation according to body weight, tube voltage,

tube current and IR strength were investigated. In chapter 5 the parameters
age, kidney function and the need for repetitive scanning (as in oncological
follow-up for example) were explored. These patient characteristics determine
whether the risk of CM or radiation are most relevant. Because ATVS choses a
particular tube voltage based on a desired contrast-to-noise ratio (CNR), and
because CNR is determined by both contrast and radiation dose, decreasing
one and increasing the other – or vice versa – will maintain a constant CNR. In
an experimental set up with 6 Göttingen minipigs optimizing either radiation
(-26 %) or CM dose (-16 %) led to comparable objective and subjective image
quality. The results suggest that it is feasible to optimize either the radiation
or the CM dose based on individual risk assessment.
10
Chapter 6 details a prospective study in 218 patients to evaluate whether
administering CM at room temperature resulted in comparable image quality,
safety and participant comfort compared to pre-warmed CM administration
(37° C [99° F]) in abdominal CT imaging. In contrast to the European guideline,
which states that pre-warming CM improves patient comfort and reduces the
number of adverse events, the present study showed that applying CM at room
temperature is noninferior with respect to image quality, safety and comfort
in this setting.

The editorial in chapter 7, outlines the 10-to-10 rule. This rule states that a 10
kV reduction in tube voltage should result in a 10 % decrease in IDR in vascular

161

Vol_BM_productie.indd 161 02/02/2022 15:23:20

 Chapter 10

studies and a 10 % decrease in dosing factor (in g I/kg) in parenchymal studies,

and vice versa. Both statements have been prospectively verified in previous
trials by our group, and one is included in this thesis (chapter 3).

In short, the present thesis proposes several parameters to base scan and
CM injection protocols on in abdominal CT imaging. It may not be easy to
incorporate all these different facets within an efficient workflow, but artificial
intelligence (AI) may provide a solution. Chapter 8 discusses the possibilities
of AI in cardiac imaging. The chapter focuses mainly on CT angiography, but the
large majority of suggestions are most likely applicable to parenchymal imaging.
For example, AI could provide an automatic attenuation check, followed by
pathology detection. When enhancement of the targeted structure is found
to be insufficient, improved scan and CM protocols could be automatically
proposed. As for the pathology check, a warning signal could be generated
whenever an acute pathology is detected. Thus AI may be helpful on different
levels to improve daily clinical workflow efficiency.

162

Vol_BM_productie.indd 162 02/02/2022 15:23:20

 Nederlandse samenvatting

Nederlandse Samenvatting

Sinds de ontdekking van de CT-scanner in 1971 zijn de contrast media (CM)

protocollen, software en de scanners zelf snel ontwikkeld. In het begin werd
er een one-size-fits all methode gebruikt. De stralingsdosis was gelijk in iedere
patiënt, net als de hoeveelheid CM die werd toegediend. Echter, door de
technische ontwikkeling van de scanner, werd het mogelijk om scans te verrichten
met een verschillende buisstroom (mAs) en buisspanning (kV). Automated
tube current modulation (ATCM) en automated tube voltage selection (ATVS)
technieken optimaliseren de stralingsdosis gebaseerd op een beeldkwaliteit
die vooraf door de gebruiker is vastgesteld, als ook op de individuele patiënt
karakteristieken. Eerdere studies hebben daarnaast aangetoond dat voor de
optimalisatie van CM protocollen in vasculaire studies de iodine delivery rate (IDR)
de meest belangrijke factor is, terwijl voor parenchymateuze studies het totale
CM volume het meest essentieel is. Vanwege de 33 keV-edge van jodium, is het
zo dat een daling in het kV – waarbij de totale stralingsdosis wordt verlaagd –
resulteert in een toename in de aankleuring van het jodium. Een verlaging van
de buisspanning kan dus een voordeel zijn voor zowel de stralingsdosis als het
CM volume. Het nadeel van het verlagen van de buisspanning is echter een
toename in ruis. Dit toont het fragiele evenwicht van het “as low as reasonably
achievable” (ALARA) principe. Het doel van de huidige thesis is om deze balans
verder uit te diepen en te onderzoeken hoe zowel de straling als CM dosis
geïndividualiseerd kan worden, gebaseerd op zowel de klinische vraag als de
patiënt karakteristieken. Met hierbij als doel een optimale beeldkwaliteit te
behalen in iedere patiënt, elke keer. 10

In hoofdstuk 2 wordt een CM injectie protocol aangepast aan het

lichaamsgewicht van de patiënt in abdominale CT geïntroduceerd. Zoals eerder
aangetoond in beeldvorming van het hart en de pulmonaal arteriën, resulteert
een CM protocol gebaseerd op lichaamsgewicht in meer homogene aankleuring
van het doelorgaan in vergelijking met een one-size-fits all protocol. Tegelijkertijd
wordt in deze eerdere studies een vermindering in de totale hoeveelheid CM
behaald in een groot deel van de populatie. Resultaten van dit hoofdstuk zijn in
lijn met deze eerdere studies: Een CM protocol gebaseerd op lichaamsgewicht
resulteert in een homogene aankleuring van de lever, in vergelijking met een
protocol waarbij iedere patiënt dezelfde hoeveelheid CM heeft gekregen.

163

Vol_BM_productie.indd 163 02/02/2022 15:23:20

 Chapter 10

De hypothese voor hoofdstuk 3 was dat een CM injectie protocol gebaseerd

op lichaamsgewicht, aangepast aan de buisspanning, resulteert in homogene
aankleuring van het lever parenchym tussen patiënten. In deze dubbelblinde,
gerandomiseerde studie werden 256 geïncludeerd in een van de vier groepen.
In groep 1 werd er gescand met de ‘gouden standaard’: 120 kV en een CM
protocol aangepast aan het lichaamsgewicht van 0.521 g I/kg. In groep 2 werd
de buisspanning verlaagd naar 90 kV, terwijl het CM protocol identiek bleef aan
dat gebruikt in groep 1. In groep 3 werd de buisspanning verlaagd naar 100 kV
en daarom werd het contrast ook verlaagd met 20 % (in vergelijking met groep
1) tot 0.417 g I/kg. In groep 4 werden patiënten gescand met 90 kV en 0.365 g
I/kg (een reductie van 30 %). De hypothese werd bevestigd dat een CM injectie
protocol gebaseerd op lichaamsgewicht en de gebruikte buisspanning leidde tot
homogene aankleuring van de lever in abdominale CT in portaal veneuze fase.

Hoofdstuk 4 onderzocht de optimale iteratieve reconstructie (IR) sterkte en

buisstroom in abdominale CT met het gebruik van reconstructie software. Deze
studie toonde door paarsgewijze vergelijking in dezelfde patiënt aan dat IR-
sterkte 4 de beste subjectieve beeldkwaliteit gaf, terwijl een 10 – 40 % verlaging
in de buisstroom mogelijk was zonder afbreuk te doen aan de objectieve en
subjectieve beeldkwaliteit.

Lichaamsgewicht, buisspanning, buisstroom en IR sterkte werden onderzocht

in de voorgaande hoofdstukken. In hoofdstuk 5 werden de parameters
leeftijd, nierfunctie en de noodzaak om een patiënt herhaaldelijk te scannen
(e.g. oncologische follow-up) meegenomen als factoren om het CM protocol op
te baseren. ATVS kiest een bepaalde buisspanning, gebaseerd op een gewenste
contrast-to-noise ratio (CNR). Echter, de CNR wordt bepaald door zowel het
contrast als de stralingsdosis. Door de een te verlagen en de ander te verhogen
– of omgekeerd – kan de CNR constant worden gehouden. In een experimentele
opzet in 6 Göttingen mini-varkentjes leidde het optimaliseren van enkel de
stralingsdosis (-26 %) of de CM dosis (-16 %) tot een vergelijkbare objectieve en
subjectieve beeldkwaliteit. Daarom kan redelijkerwijs worden aangenomen dat
het mogelijk is enkel de stralings- of de CM dosis te optimaliseren, afhankelijk
van een individuele risicobepaling.

164

Vol_BM_productie.indd 164 02/02/2022 15:23:20

 Nederlandse samenvatting

Hoofdstuk 6 onderzocht prospectief in 218 patiënten of het toedienen van CM

op kamertemperatuur resulteerde in vergelijkbare beeldkwaliteit, veiligheid en
patiënt comfort ten opzichte van verwarmd CM (37° C [99° F]) in abdominale CT.
In tegenstelling tot de Europese richtlijn, die stelt dat CM verwarming de patiënt
meer comfortabel maakt en de kans op ongewenste voorvallen verminderd,
toont de huidige studie dat CM op kamertemperatuur non-inferieur was met
betrekking tot beeldkwaliteit, veiligheid en comfort.

In hoofdstuk 7 wordt de 10-tot-10 regel samengevat in een editorial. Deze

regelt stelt dat een 10 kV verlaging in buisspanning zou moeten leiden tot een
10 % vermindering in IDR voor vasculaire studies en 10 % verlaging van het
aantal g I/kg in CT’s van de parenchymateuze organen en omgekeerd. Beide
statements zijn prospectief onderbouwd in eerdere studies verricht door onze
groep of gepresenteerd in de huidige thesis.

Deze thesis zet verschillende parameters uiteen waarop het scan en CM

injectie protocol in abdominale CT gebaseerd kan worden. Om al deze
verschillende facetten te integreren en tegelijkertijd de workflow efficiënt te
houden, zou artificiële intelligentie (AI) een oplossing kunnen zijn. Hoofdstuk 8
bediscussieert de verschillende mogelijkheden van AI in cardiale beeldvorming.
Alhoewel gericht op het hart, zouden deze suggesties ook van toepassing
kunnen zijn op de abdominale beeldvorming. AI zou bijvoorbeeld kunnen
zorgen voor een automatische aankleurings-check, gevolgd door de detectie
van eventueel aanwezige acute pathologie. In het geval van onvoldoende
aankleuring van het doelorgaan kan AI een verbeterd scan en CM protocol 10
voorstellen en de patiënt opnieuw laten scannen, voordat deze terug gaat naar
huis. Met betrekking tot de pathologiecontrole kan een waarschuwingssignaal
afgaan bij de aanwezigheid van acute pathologie. Door AI te introduceren zou
op deze manier de dagelijkse klinische workflow efficiënter kunnen worden.

165

Vol_BM_productie.indd 165 02/02/2022 15:23:20

Vol_BM_productie.indd 166 02/02/2022 15:23:21
 CHAPTER
11
Scientific impact

Vol_BM_productie.indd 167 02/02/2022 15:23:22

 Chapter 11

11.1 Research

Radiation dose and contrast media (CM) together ensure image quality in
(abdominal) computed tomography (CT) imaging. However, using a one size
fits all protocol may not be the best tactic. Optimizing radiation and CM dose
will result in individualized scan and CM protocols in which, ideally, each patient
will receive the optimal amount of both to reach diagnostic image quality. In
such optimization, however, the type of CT study performed must be taken into
account. In vascular studies the iodine delivery rate (IDR, in gI/s) is considered
the most decisive factor. For parenchymal studies, the CM volume (in ml) is the
most important parameter to reach optimal enhancement of the target organ
(1). The aim of the current thesis was to find the optimal radiation and CM dose
for each patient in abdominal imaging. This thesis proposes a 10-to-10 rule of
thumb to individualize scan and CM injection protocols. A 10 kV decrease in
tube voltage should be accompanied by a 10 % decrease in IDR for vascular
studies and a 10 % decrease in dosing factor for a parenchymal CT, and vice
versa. Results of a randomized controlled trial (COMpLEx trial) confirmed this
easy to implement rule of thumb in abdominal imaging.

To date a disagreement exists between the European and American

guidelines regarding the necessity to pre-warm CM to body temperature
before intravenous administration (2, 3). The current thesis provides high
level evidence that pre-warming CM does not result in increased image quality,
safety, or patient comfort in abdominal CT imaging (CATCHY trial). This result
could improve work flow efficiency in daily clinical practice, as there may be no
need to store CM in a warming cabinet.

Apart from body weight, tube voltage and CM temperature, this thesis proposes
to add the patient characteristics age and kidney function to the parameters
used for protocol optimization. In younger patients and patients in need of
frequent scanning, radiation dose reduction is preferred, whereas patients with
reduced kidney function (more frequently seen in the elderly population) may
benefit more from CM dose reduction. These additional patient characteristics
were evaluated in an animal feasibility study with promising results.

168

Vol_BM_productie.indd 168 02/02/2022 15:23:22

 Scientific impact

The present thesis tackles different parameters step by step. However,

combining all pertinent factors into one protocol is challenging. Artificial
intelligence (AI) could be the solution to further optimize daily workflow in
the CT department. AI could for example, aid in radiation and CM protocol
individualization, detect insufficient image quality at an early stage, and provide
a warning signal upon the detection of an acute pathology requiring immediate
viewing by a radiologist.

11.2 Relevance

In the past, a one-size fits all protocol was used for both radiation – tube
voltage and tube current – and CM dose protocols. However, rapid technical
developments made it possible to adapt tube current and tube voltage to the
clinical question and patient body weight, substantially reducing radiation dose
(4). In daily clinical routine worldwide, CM is still often administered in a one-
size-fits all fashion. Previous studies from our group and of the present thesis
show that individualizing the CM protocol based on body weight results in
more homogeneous enhancement in cardiac, pulmonary artery, and abdominal
imaging. Furthermore, a simultaneous reduction in total injected CM volume
was achieved in a large percentage of the population (5, 6).

Radiation and CM dose are often treated as two separate entities. However,
considering both parameters in conjunction opens new doors. The 10-to-10
rule offers an easy-to-use and readily implemented rule of thumb to adapt both
parameters to one another. By introducing an opportunity to adapt either the
radiation or the CM dose – depending on age, kidney function or the necessity
for repetitive scanning – protocols can be further optimized based on individual 11
risk assessment.

The present thesis provides a manual on how to individualize CT protocols in

daily clinical practice. By applying the proposed rules, it is possible to reach
sufficient image quality in each patient, every time, whilst maintaining the
perfect balance between radiation and CM dose.

169

Vol_BM_productie.indd 169 02/02/2022 15:23:22

 Chapter 11

11.3 Target groups

There are four groups for which this thesis could be relevant.

1. Radiologists

Radiologists in general are aware of the fact that radiation and CM protocols
influence image quality. However, there is too little awareness of how protocols
influence important aspects such as lesion characterization. Different protocols
– between hospitals, scanners, and moments in time – will result in different
attenuation levels. In kidney lesions for example, attenuation predicts the
likelihood of a malignant lesion. Therefore, homogeneity between and within
patients is desirable in order to draw reliable conclusions from each scan.
The current thesis provides an easy-to-use rule of thumb to reach such
homogeneous enhancement in both vascular and abdominal CT imaging.

2. Radiologic Technologists

In the Netherlands, technicians are responsible for acquiring the scan according
to protocols as determined by the radiologist. The information in the present
thesis may give technician’s more insight into why and how protocols are
optimized. Furthermore, the suggestions made with regard to the introduction
of AI may simplify their job.

3. Referring physicians

Clinicians are happy with a performed CT scan when it is easy to assess and
has diagnostic image quality. In order to make sure that we can provide that
‘pretty’ CT scan, clinicians have to provide a scan indication, clinical background,
and correct patient body weight and kidney function. While the current thesis
may be too focused on the technical aspects of CT to capture the clinicians’
imagination, a little glimpse into the world of the CT department would help
them understand why these particular questions are asked of them.

170

Vol_BM_productie.indd 170 02/02/2022 15:23:22

 Scientific impact

4. The patient

Providing the patient with a CT scan with the highest achievable image quality,
assists in diagnosing a diversity of diseases. In addition, decreasing radiation
and CM dose diminishes the associated life time attributable cancer risk and
possible drop in renal function. Last but not least, comfort is important to the
patient. This is reflected in the fact that it was quite easy to find patients willing
to participate in the CATCHY trial.

11.4 Activity

Most patients are somewhat familiar with X-ray imaging, but CT and Magnetic
Resonance Imaging (MRI) are often confused with one another. In addition,
most people do not really know what the work of a radiologist entails. At a first
glance this shouldn’t be a problem, but it may be beneficial for both referring
physician and patient to know a little more about radiology. Referring physicians
are informed through presentations. Patients may be reached through social
media. Creating awareness of what is done in the CT department to reach
diagnostic image quality may improve patients’ understanding of procedures,
and perhaps even the existence of waiting lists.

References

1. Bae KT. Intravenous contrast medium administration and scan timing at CT:
considerations and approaches. Radiology. 2010;256(1):32-61.
2. American College of Radiology. Manual On Contrast Media: 2021 [Available from:
https://www.acr.org/-/media/ACR/Files/Clinical-Resources/Contrast_Media.pdf.
11
3. European Society of Urogenital Radiology. ESUR guidelines on contrast agents
European Society of Urogenital Radiology 10.0 2018 [Available from: http://www.
esur.org/fileadmin/content/2019/ESUR_Guidelines_10.0_Final_Version.pdf.
4. Lell MM, Wildberger JE, Alkadhi H, Damilakis J, Kachelriess M. Evolution in computed
tomography: the battle for speed and dose. Invest Radiol. 2015;50(9):629-44.
5. Mihl C, Kok M, Altintas S, Kietselaer BL, Turek J, Wildberger JE, et al. Evaluation
of individually body weight adapted contrast media injection in coronary CT-
angiography. Eur J Radiol. 2016;85(4):830-6.
6. Hendriks BM, Kok M, Mihl C, Bekkers SC, Wildberger JE, Das M. Individually
tailored contrast enhancement in CT pulmonary angiography. Br J Radiol.
2016;89(1061):20150850.

171

Vol_BM_productie.indd 171 02/02/2022 15:23:22

Vol_BM_productie.indd 172 02/02/2022 15:23:24
 CHAPTER
12
Dankwoord
Curriculum Vitae
List of Publications

Vol_BM_productie.indd 173 02/02/2022 15:23:25

 Chapter 12

Als ‘materiaal en methoden’ goed in elkaar zitten, dan leidt dit vanzelf tot een
goede studie. Deze stelling is wat mij betreft van toepassing op de gehele
totstandkoming van dit boekje, zonder een solide basis (onderzoeksteam,
afdeling beeldvorming, familie en vrienden) had deze thesis niet tot stand
kunnen komen en zeker niet tot het plezier geleid dat ik er nu aan heb gehad.
Graag zou ik een aantal personen in het bijzonder willen bedanken:

Prof. Dr. J.E. Wildberger, beste Joachim, uren hebben we samen zitten
brainstormen over de opzet van de verschillende studies. Dan was ik overtuigd
van een bepaalde richting, maar kon je het plan met één kritische vraag
onderuithalen, uiteindelijk leidend tot een proefschrift waarbij geen moment
is verspeeld. Ik wil je enorm bedanken voor de tijd die je in je drukke agenda
hebt vrijgemaakt om mij te begeleiden, voor je interesse ook in mij als persoon
en voor alle mooie kansen en de prettige samenwerking, waar ik veel van heb
geleerd en hopelijk nog lang van mag blijven leren.

Dr. C. Mihl, beste Casper, onze eerste kennismaking was, zoals ik me kan
herinneren, op jouw verjaardagsfeest in het Forum, waarop je een gele eendjes
onesie droeg en waar ik via verschillende wederzijdse vrienden mee naartoe
werd gesleurd als AIOS in spe. Wat mij betreft een vrij typerend beeld voor
onze latere samenwerking. Je bent altijd beschikbaar, geeft bizar snelle, to-the-
point, eerlijke feedback en dat alles met een grote glimlach. Daarnaast wil ik je
bedanken voor al je connecties waar ik op mee heb kunnen liften. Ik had me
geen prettigere samenwerking kunnen wensen en hoop dat we hier nog lang
mee door kunnen gaan. We hebben een biertje verdiend!

Dr. E.C. Nijssen, beste Estelle, een aantal papers was zeker niet zo gemakkelijk
gepubliceerd geweest als ze niet eerst door de Estelle-check waren geweest.
Bewonderenswaardig hoe je de goede vragen kunt stellen over een onderwerp
wat niet echt dicht bij dat van jezelf ligt. Je kritische feedback, met altijd een
vriendelijke noot zijn super leerzaam geweest en ik hoop in de toekomst nog
vaker van je expertise gebruik te mogen maken.

Babs, bedankt dat je mijn research-buddy, mijn vraagbaak om voor de zoveelste

keer uit te leggen hoe (onder andere) een ‘collimator’ ook alweer werkt en mijn
afgetrainde body-double (hoe vaak zijn we niet door elkaar gehaald!?) bent, voor

174

Vol_BM_productie.indd 174 02/02/2022 15:23:25

 Dankwoord

gezelligheid op feestjes, borrels en congressen zorgt en altijd in bent voor een

cappuccino. Samen een fantoom opschuren tot er barstjes in komen, je vent
uitlenen voor een extra Engelse check, gesandwiched in de scanner voor een
promotiefilmpje, party-en op festivals en als kers op de taart ben je een van
mijn paranimfen. Dat we nog maar lang veel plezier mogen beleven!

Nienke, speciaal voor jou een prominent plaatsje in dit dankwoord. Je hebt me
enorm op weg geholpen in de struggles van het onderzoek: hoe te beginnen,
wie en wat moet ik waar vinden, harde schijven, kluisjes-sleutels, PhD tracks
en de eerste opzet van een artikel. Daarnaast moet ik tegelijkertijd met jou
Youtube bedanken. Door jou heb ik geleerd dat dit een onmisbare bron is voor
‘waterdichte’ statistiek!

Lieve CT-laboranten, ik hoop dat jullie beseffen dat dit boekje er zonder jullie
hulp niet was geweest. Als ik weer over een nieuwe studie kwam vertellen
zag ik jullie soms wat moeilijk kijken, maar vol goede moed was het varkentje
uiteindelijk altijd binnen no-time gewassen. Jullie gedrevenheid, enthousiasme
en professionaliteit hebben ervoor gezorgd dat we steeds supersnel patiënten
hebben kunnen includeren en dat vrijwel alle data gebruikt kon worden. Graag
zou ik drie personen specifiek uitlichten: Ankie, bedankt dat je altijd flexibel en
enthousiast bent, mede door jou als teamleider is CT altijd een prettige plek om
te zijn. Serena, je gaat het zeker weten fantastisch doen! Dank voor al het extra
werk dat je voor me hebt gedaan! Jef, bedankt dat je veel slimmer bent dan ik!

Prof. Dr. M.W. de Haan, het heeft even geduurd, maar het lukt me inmiddels
al een tijdje om je gewoon Michiel te noemen. Mijn eerste echte gesprek bij de
radiologie was met jou en daarmee is de cirkel met jou als voorzitter van de
leescommissie rond, bedankt! De overige leden van de beoordelingscommissie,
Prof. Dr. Steven Oldedamink, Prof. Dr. Roger Rennenberg, Dr. Nils Planken
en Dr. Doenja Lambregts wil ik eveneens hartelijk danken voor de grondige
beoordeling van mijn proefschrift.

Dames en gepensioneerde-heer van de administratie en Reza, zonder jullie

hadden patiënten niets geweten van mijn studies en was dit proefschrift er
dus ook niet geweest. Reza (en Marcel) dank voor het uitprinten van de enorme
lading folders en je altijd gulle glimlach en interesse. Bea, door jou kwam er een

175

Vol_BM_productie.indd 175 02/02/2022 15:23:25

 Chapter 12

heuse administratie op gang om de patiënten voor de verschillende studies te

bereiken, dank voor je stapje extra.

Alle co-auteurs, van iedereen afzonderlijk heb ik iets geleerd, waarvoor dank.
Specifiek zou ik graag Dr. Sander van Kuijk bedanken, die de kracht bezit om een
oplossing te bedenken voor elk statisch raadsel dat in mijn hoofd onoplosbaar
was geworden.

Dames van het secretariaat, over een solide basis gesproken, in de afgelopen
jaren zijn er zoveel kleine tot grote problemen waar jullie me mee hebben
geholpen: vergeten CAT’s, vergaderruimtes, sleutels, laptops, taart, afspraken,
agenda’s, contacten, de weg, enveloppen en ik ben er vast nog een paar
vergeten. Duizendmaal dank.

Dr. A.A. Postma-Jacobi, beste Linda, als opleider heb je ervoor gezorgd dat ik
mijn opleiding volledig zoals ik wilde heb kunnen vormgeven. Daarnaast ben
je een van de personen die het eerste research-zaadje bij me heeft gepland,
wie had gedacht dat een presentatie over de fameuze sniff-test dat teweeg zou
kunnen brengen. Dankjewel voor al je enthousiasme, je oprechte interesse en
je luisterend oor.

Doenja en Max, jullie zijn toch wel een klein beetje mijn voorbeeld geweest: Wat
zij doen, dat wil ik ook! Daarnaast was ik zonder Doenja’s lichte dwang wellicht
nooit tot het inzicht gekomen deze opleiding te gaan doen. Bedankt voor al
jullie adviezen, maar ook de vele drankjes en dansjes.

Prof. Dr. M. Das, Marco, jij was degene die vroeg of ik al eens nagedacht had
over research binnen jouw team. Het technische onderwerp was niet hetgeen
dat me direct aansprak, maar het team dat je rondom je had verzameld wel.
Zelf ben je helaas snel na mijn start vertrokken, maar toch hartelijk dank voor
het gegeven vertrouwen.

Alle assistenten en stafleden van de radiologie en nucleaire geneeskunde

bedankt voor de prettige werkplek. Specifiek de abdomen en cardio groep
bedankt voor de flexibiliteit wanneer ik afspraken had tijdens klinische taken.
Anna, Nicky en Maud bedankt voor al jullie gezelligheid, borrels, (super originele

176

Vol_BM_productie.indd 176 02/02/2022 15:23:25

 Dankwoord

en goed bedachte) cadeautjes, steun, betrokkenheid en heerlijke etentjes, dat

er nog maar veel mooie momenten mogen volgen!

Michelle, Maastricht zou echt veel minder leuk zijn als jij er niet zou wonen! Bij
ons blijft het nooit bij één drankje, wijdt het aan ons gebrek aan ruggengraat,
maar ik zou dat ruggengraat voor geen goud willen terugvinden. Super bijzonder
om je beste vriendin op slechts 650 meter afstand te hebben wonen. Dank voor
alle etentjes, drankjes, vakanties, spelletjes, het luisteren, voor je oprechtheid,
kookkunsten, rij- en fietslessen, dansjes, en het feit dat je mij soms minder
serieus maakt. Het is niet meer dan logisch dat jij een van mijn paranimfen bent!

Papa en Mama, ik weet dat jullie het raar vinden dat je hier bedankt wordt,
maar het hoort nou eenmaal zo. Daarnaast hebben alle uurtjes wiskunde,
natuurkunde en scheikunde oefenen, in combinatie met grote hoeveelheid
snoep en het verder he-le-maal niks in huis hoeven doen, er wel voor gezorgd
dat het gelukt is. Dankjewel dat jullie me hebben geleerd dat ik alles mag
zeggen (als ik het maar netjes doe), dat jullie op alle reis- en studieplannen
altijd enthousiast hebben gereageerd, jullie eerlijkheid, voor de hulp bij alle
verhuizingen en dat ik weet dat ik jullie altijd mag bellen als er iets is. Broeder,
Sjimmie, jij hebt echt niks gedaan aan dit proefschrift, maar toch bedankt ;),
biertje?

Lieve Mickel, dankjewel dat ik van jou alles mag en kan doen wat ik leuk vind.
Dat je het nooit zegt als je liever had gehad dat ik geen avond, week of maand
weg ga en me juist stimuleert om dat soort uitdagingen aan te gaan. Dankjewel
dat je snapt dat mijn primaire levensbehoeftes (honger, dorst, moe en koud)
gewoon heel erg belangrijk zijn en vaak herhaald moeten worden. Dankjewel
dat je de perfecte bliksemafleider bent als ik op andere plekken alle vrolijkheid
al heb vergeven. En uiteraard bedankt voor je onmiskenbare photoshop talent
;). Dat we nog maar veel avonturen mogen beleven!

177

Vol_BM_productie.indd 177 02/02/2022 15:23:25

Vol_BM_productie.indd 178 02/02/2022 15:23:26
 CHAPTER
12
Dankwoord
Curriculum Vitae
List of Publications

Vol_BM_productie.indd 179 02/02/2022 15:23:27

 Chapter 12

180

Vol_BM_productie.indd 180 02/02/2022 15:23:27

 Curriculum Vitae

Bibi Martens was born on the 4 th of December 1986 in Roosendaal, the

Netherlands. In 2005 she graduated from the Strabrecht College in Geldrop,
The Netherlands. In the same year, she started at the Faculty of Psychology and
Neuroscience in Maastricht with her Bachelor in Psychology, with a 6-month
elective internship at James Cook University in Smithfield (Australia). She
received her bachelor degree in 2010 and started her medical training at the
Faculty of Health, Medicine and Life Sciences at the University of Maastricht
in 2008. After obtaining the medical degree in 2014 she started her radiology
residency in 2015 at Maastricht University Medical Center (MUMC), under the
supervision of dr. A.A. Postma-Jacobi. In 2016 she commenced her PhD project
under the supervision of Prof. dr. J.E. Wildberger, dr. C. Mihl and dr. E.C. Nijssen
with as a main goal the optimalization of both radiation and contrast media
dose in abdominal computed tomography. The results led to present thesis
and were presented at several (inter)national congresses. She was awarded
for parts of the research with a ‘Student Travel Award, for one of the best
abstracts during the RSNA annual meeting in Chicago and the ‘René Vogels
Reisstipendium’, which was not used due to the COVID-19 pandemic. She
finished her radiology training in January 2021, after which she started as a
fellow in cardiovascular imaging at the MUMC. She remains involved in this
research topic.

181

Vol_BM_productie.indd 181 02/02/2022 15:23:27

Vol_BM_productie.indd 182 02/02/2022 15:23:29
 CHAPTER
12
Dankwoord
Curriculum Vitae
List of Publications

Vol_BM_productie.indd 183 02/02/2022 15:23:29

 Chapter 12

This thesis

Martens B, Hendriks BMF, Eijsvoogel NG, Wildberger JE, Mihl C. Individually

body weight–adapted contrast media application in computed tomography
imaging of the liver at 90 kVp. Invest. Rad. 2019;54(3):177-82.

Martens B, Wildberger JE, Hendriks BMF, Van Kuijk SMJ, Nijssen EC, Peters
NHGM, De Vos-Geelen J, Mihl C. A solution for homogeneous liver enhancement
in computed tomography: results from the COMpLEx trial. Invest Radiol.
2020;55(10):666-72.

Martens B, Bosschee JGA, Van Kuijk SMJ, Jeukens CRLPN, Brauer MTH,
Wildberger JE, Mihl C. Finding the optimal tube current and iterative
reconstruction strength in liver imaging; two needles in one haystack. Under
review.

Martens B, Jost G, Mihl C, Wildberger JE, Schmidt B, Flohr T, Pietsch H.

Individualized scan protocols in abdominal computed tomography: radiation
versus contrast media dose optimization. Investigative Radiology 2021; Epub
ahead of print.

Martens B, Wildberger JE, Van Kuijk SMJ, De Vos – Geelen J, Jeukens CRLPN, Mihl
C. Influence of contrast material temperature on patient comfort and image
quality in computed tomography of the abdomen (CATCHY): a randomized
controlled trial. Invest Radiol. 2021. 2022;57(2):85-89.

Martens B*, Hendriks BMF*, Mihl C, Wildberger JE. Tailoring contrast media
protocols to varying tube voltages in vascular and parenchymal CT imaging: the
10-to-10 rule. Invest Radiol. 2020;55(10):673-6. * shared first authorship

Martens B, Hendriks BMF, Wildberger JE, Mihl C. Book chapter: Artificial

intelligence-based contrast medium optimization. Carlo N. De Cecco, Marly
van Assen, Tim Leiner (ed): Artificial Intelligence in Cardiothoracic Imaging - ISBN:
978-3-030-92086-9 Springer, 1st edition 2022.

184

Vol_BM_productie.indd 184 02/02/2022 15:23:30

 List of publications

Other publications

Eijsvoogel NG, Hendriks BMF, Willigers JL, Martens B, Carati LF, Horehledova
B, et al. Personalization of injection protocols to the individual patient’s blood
volume and automated tube voltage selection (ATVS) in coronary CTA. PLoS
One. 2018;13(9):e0203682.

Hendriks BMF, Eijsvoogel NG, Kok M, Martens B, Wildberger JE, Das

M. Optimizing pulmonary embolism computed tomography in the age of
individualized medicine; a prospective clinical study. Invest. Rad. 2018;53(5):306-
12.

Eijsvoogel NG, Hendriks BMF, Nelemans PJ, Mihl C, Willigers J, Martens B,

Wildberger JE, Das M. Personalization of CM injection protocols in coronary
computed tomographic angiography (People CT Trial). Contrast Media &
Molecular Imaging 2020(7):1-12.

Eijsvoogel NG, Hendriks BMF, Martens B, Gerretsen SC, Gommers S, van Kuijk
SMJ, Mihl C, Wildberger JE, Das M. The performance of non-ECG gated chest CT
for cardiac assessment - The cardiac pathologies in chest CT (CaPaCT) study.
Eur J Radiol. 2020;130:109151.

Mihl C, Martens B. Book chapter: All about CT contrast agents. Gaemperli O,

Pontone G, Nieman K, Maurovich-Horvat P (ed): EACVI Handbook of Cardiac
CT (in preparation).

Hendriks BMF, Martens B, Mihl C. Pre-procedural computed tomography

in transcatheter pulmonary valve replacement: The first steps towards
standardization of image quality. IJC Heart & Vasculature. 2020;29:100542.

Kemper CA, Mihl C, Martens B, McDermott MM, Hendriks BMF. Performance

of centargo: a novel piston-based injection system for high throughput in CE
CT. Submitted to Journal of Medical Devices: Evidence and Research.

Martens B, Driessen RGH, Brandts L, Hoitinga P, Van Veen F, Driessen M,

Weberndörfer V, Kietselaer B, Ghossein-Doha C, Gietema HA, MaastrICCht

185

Vol_BM_productie.indd 185 02/02/2022 15:23:30

 Chapter 12

Collaborators, Vernooy K, Van der Horst ICC, Wildberger JE, Van Bussel BCT,
Mihl C. Coronary artery calcifications are associated with a worse development
of multi-organ failure in patients with a severe COVID-19 infection; longitudinal
results of the Maastricht Intensive Care COVID cohort. Under review at The
Journal of Thoracic Imaging.

186

Vol_BM_productie.indd 186 02/02/2022 15:23:30

Vol_BM_productie.indd 187 02/02/2022 15:23:30
Vol_BM_productie.indd 188 02/02/2022 15:23:30