Advice From STUK September 2012 Paediatric CT

ADVICE FROM STUK / SEPTEMBER 2012
Guidelines for paediatric

CT examinations
Säteilyturvakeskus
Strålsäkerhetscentralen
Radiation and Nuclear Safety Authority
Translation. Original text in Finnish.
Several paediatric radiologists and medical physics experts who are specialized in paediatric
CT imaging and also scanner suppliers have taken part in preparing this guide. STUK expres-
ses its greatest thanks to them for sharing their expertise and their good cooperation.
Katja Merimaa, STUK - Radiation and Nuclear Safety Authority (ed.)
Raija Seuri, Helsinki University Hospital (ed.)
Ritva Bly, STUK - Radiation and Nuclear Safety Authority
Anna Föhr, Helsinki University Hospital
Eero Hippeläinen, Helsinki University Hospital
Minna Husso, Kuopio University Hospital
Touko Kaasalainen, Helsinki University Hospital
Eero Kauppinen, Central Finland Central Hospital
Reetta Kivisaari, Helsinki University Hospital
Mika Kortesniemi, Helsinki University Hospital
Kirsi Lauerma, Helsinki University Hospital
Laura Martelius, Helsinki University Hospital
Marja Perhomaa, Oulu University Hospital
Jukka Schildt, Helsinki University Hospital
Sanna Toiviainen-Salo, Helsinki University Hospital
Erja Tyrväinen, Kuopio University Hospital
Leena Valanne, Helsinki University Hospital
Representatives of the scenner suppliers: GE, Philips, Siemens and Toshiba
Preparations of the guide was also supported by the following persons:
Atte Karppinen, Helsinki University Hospital
Antti Markkola, Helsinki University Hospital
Aki Ikonen, Central Finland Central Hospital
Leena Valanne, Helsinki University Hospital
Miika Nieminen, Oulu University Hospital
Juha Nikkinen, Oulu University Hospital
The members of the Paediatric Radiologist Club of the Radiological Society of Finland.
ISSN-L 1799-9510
ISSN 1799-9510 • ISBN 978-952-478-791-8 (pdf)
Contents
1. Introduction................................................................................................................ 5
2. General information................................................................................................... 6
3. Basics of optimizing CT examinations ....................................................................... 7
3.1 Scanning parameters.......................................................................................... 7
3.2 Dose and image quality..................................................................................... 11
3.3 Using radiation shielding................................................................................... 14
3.4 Preparation for the examination........................................................................ 15
3.5 Literature........................................................................................................... 17
4. Vendor specific features........................................................................................... 18
4.1 General Electric (GE)......................................................................................... 18
4.2 Philips................................................................................................................ 19
4.3 Siemens............................................................................................................ 21
4.4 Toshiba.............................................................................................................. 22
5. Head region.............................................................................................................. 24
5.1 Head.................................................................................................................. 24
5.2 Head scanning in the assessment of ventriclular size ...................................... 24
5.3 Scanning the cranial sutures............................................................................. 25
5.4 Literature........................................................................................................... 25
6. Ear, nose and throat region...................................................................................... 26
6.1 Paranasal sinuses.............................................................................................. 26
6.2 Ear region.......................................................................................................... 26
6.3 Facial bones....................................................................................................... 27
6.4 Literature........................................................................................................... 27
7. Neck region.............................................................................................................. 28
8. Chest region............................................................................................................. 29
8.1 Thorax................................................................................................................ 29
8.2 High-resolution CT............................................................................................. 30
8.3 Chest CT-angiography examinations ................................................................. 33
8.4 Literature........................................................................................................... 33
9. Abdominal region..................................................................................................... 34
9.1 CT-angiography of the abdominal region........................................................... 35
9.2 Literature........................................................................................................... 35
10.Trauma..................................................................................................................... 36
10.1 Head trauma...................................................................................................... 36
10.2 Spinal trauma ................................................................................................... 36
10.3 Trauma of thorax and abdomen ........................................................................ 37
10.4 Performing trauma examination........................................................................ 37
10.5 Literature........................................................................................................... 38
3
11. Orthopaedic examinations....................................................................................... 39
11.1 Measuring leg length discrepancy from the planning image............................ 39
11.2 Joints................................................................................................................. 39
11.3 Growth plate scanning for evaluation of growth arrest..................................... 40
11.4 DDH in cast....................................................................................................... 40
11.5 Spine................................................................................................................. 40
12.Hybrid nuclear medicine examinations.................................................................... 41
12.1 Literature........................................................................................................... 42
4
Guidelines for paediatric CT
examinations
1. Introduction
The Radiation and Nuclear Safety Authority (STUK) has previously published two guidebooks in
2005 and 2008 giving guidance for paediatric x-ray procedures, including indications, guidelines
for selecting imaging parameters, optimization and proper use of radiation shields. Although
both of the previous guides briefly discussed also computer tomography (CT) examinations,
this guide reviews more closely and accurately the guidelines and taking into account the
development of CT technology and the changed scanning routines. There is also a separate
guide for determining patient doses in x-ray procedures published in 2004 that gives guidelines
for dose measurements of CT examinations and determines the dose quantities more closely.
This new guide gives the basics of optimizing CT examinations and some technical features
of the CT scanners of the four most common CT manufacturers in Finland. Justifying and
optimizing CT examinations (indications, performing the examination) are discussed covering
each anatomical area, and separately for trauma and orthopaedic examinations. In addition,
guidelines for CT examinations in nuclear medicine (the so called hybrid imaging) are given.
5
2. General information
Children are of special concern in radiation protection because radiation exposure during
childhood causes larger additional cancer risk than a corresponding exposure in adulthood.
Therefore justification and optimization of paediatric examinations needs special attention.
Although we try to avoid imaging techniques that use ionizing radiation in paediatric
examinations, CT is still needed in paediatric imaging.
A paediatric CT examination should always be planned individually for each child. Routine
examination practices without individual consideration should be avoided, and only the series
essential for diagnosis should be done. Fast multi-slice scanners have reduced the need for
sedation, and most examinations can be performed even without a breath hold. The requirements
for a successful examination include: professional performance, carefully planned working
phases and advising the child, parents and other assisting persons in a peaceful atmosphere.
Criteria for a good CT examination include
• adequately versatile, case-specifically tailored imaging practises
• minimizing the length of the scan range according to the indications and anamnesis
received
• acceptable image quality, being assessed case-specifically by the responsible radiologist (e.g.
noise level and slice thickness used on viewing) so that the patient’s radiation exposure
is kept as low as possible.
Guidance for using radiation shielding is given in item 3.3.
6
3. Basics of optimizing CT examinations
Children are smaller than adults and thus the attenuating tissue layer is also smaller, so the
radiation scatters and attenuates less when going through a child than when going through
an adult. Thus lower dose levels may give sufficient image quality.
3.1 Scanning parameters

Scanning parameters affect the patient’s radiation exposure but also the scan range and the
patient’s size affect patient’s individual dose distribution and the organs exposed to radiation.
Since the dose display values of a CT scanner don’t consider the individual patient features,
the dose display values can’t be directly used to assess the patient’s individual radiation risk
resulting from a CT scan, especially in the case of children.
Scanning voltage and tube current

A patient’s radiation dose is directly proportional to the tube loading (mAs, the product of the
tube current and scanning time) of the x-ray tube. On the other hand, image graininess (noise)
is inversely proportional to the square root of radiation dose. Thus the effect of changing the
mAs on patient dose and image noise can be predicted by calculations. The tube voltage affects
both the dose and image quality in a more complicated way.
A patient’s radiation dose is directly proportional to the tube loading (mAs). If the mAs value
is doubled, dose will also be doubled.
A lower tube voltage (80–100 kV) can be used for optimization, especially in paediatric and
contrast imaging. With a lower tube voltage the image contrast will improve and the dose will
reduce as much as tens of percents. At the same time the image noise will increase, but the
increased contrast will compensate for the increased noise, and thus the net effect to image
quality is often positive. It may take a while for a radiologist to get used to the changed image
contrast.
Planning image
The scan range is selected from the planning image (scout image or topogram). The planning
image should be long enough to cover the whole area of interest, but unnecessary long scan
ranges should be avoided. The automatic exposure control (AEC) generally uses the attenuation
information obtained from the planning image to estimate the tube current needed in different
parts of the patient (z direction). It is important not to have any extra attenuating material e.g.
radiation shields, on the selected scan area. The tube voltage should normally be the same as
for the actual scanning so that the AEC works properly.
Beam width and slice thickness

The nominal beam width in the direction of the patient’s long axis (z direction) determines
the length of the simultaneously scanned volume. Using beam collimators the beam width is
mechanically limited as desired. The nominal slice thickness (collimated slice thickness) can be
7
selected from the technical selections (e.g. 0.625 mm). The nominal slice thickness is connected to
the size of the detector elements. Nominal beam width is the nominal slice thickness multiplied
by the number of the slices (e.g. 64 · 0.625 mm = 40 mm).
The reconstructed slice thickness is the slice thickness of the viewed slices created from the
raw data and it can’t be smaller than the nominal slice thickness used for data acquisition
(the size of the detector elements used). The thinner the slices one wants to view, the higher
radiation dose is needed to obtain the desired image quality level (contrast-to-noise ratio).
The smaller the object one wants to view, the thinner reconstructed slice thickness is needed.
Usually the reconstructed slice thickness is chosen according to the indication and the child’s
size, e.g. for small children 2–3 mm and for larger 4–5 mm and for example in angiographies
even 1–2 mm may be needed.
Nominal slice thickness = slice thickness used in scanning

Reconstructed slice thickness = slice thickness used in viewing
In helical scanning some extra image data is collected at the beginning and at the end of the
scan range for image reconstruction, so the radiation exposure also expands outside the defined
scan range (the so called overranging effect). Proportion of the overranging is relatively larger
with short scan ranges, and can be up to 20% of the total dose. Therefore, in short helical series
using a narrower part of the detector can be considered (so giving a smaller nominal beam
width), thus the overranging will be smaller. However, only about 10% dose reduction can be
achived and the scan time will increase respectively; reducing the beam width to half doubles the
scan time. In some of the new scanners the helical overranging has been blocked with a special
adaptive collimation. This should be verified by the scanner supplier or a medical physics expert.
The size of the scanned and displayed field of view, SFOV and DFOV
The beam size in the scanning plane (axial plane parallel to the patient’s long axis, xy-plane)
can be altered with most of the scanners by choosing the scan field of view (SFOV), e.g. head
or body. For children a smaller SFOV should be used if possible due to their smaller size. The
patient contours should fit inside the SFOV so that the partial volume artefacts can be avoided.
The size of the display field of view (DFOV) defines the objects visible in the reconstructed
images and the final spatial resolution of the image in the axial plane. The axial CT image
consists typically of image matrix of 512 x 512 pixels. The pixels size defines the visibility of
small details, and it gets larger as DFOV is expanded. Accordingly, with a smaller DFOV the
pixel size is reduced and thus spatial resolution gets better but at the same time the noise level
increases. DFOV cannot be larger than SFOV.
CT scanners usually have one or more bowtie filters (beam shaping filters) that affect both
the image quality and dose distribution in the axial plane. They maintain the high radiation
intensity in the thickest and most attenuating part, in the middle of the patient, and lower
radiation intensity on the surface of the patient. Thus a bowtie filter evens the dose and noise
distribution in the axial plane, if the patient is centered properly. Centering the patient carefully
to the isocenter assures the proper performance of the bowtie filters. The shape of the bowtie
filter is defined according to the selected SFOV in the scanning program (if different filters
are available).
8
Automatic exposure control (AEC)
Automatic exposure control (mA modulation, AEC) compensates for beam attenuation at
different parts of the patient by changing the tube current in order to maintain constant
image quality over the whole scan range. The tube current can be changed both while the tube
is rotating around the patient and as the table moves in the direction of patient’s long axis
(z-modulation). The AEC software predicts the patient’s attenuation either beforehand from
the planning images or real time during the scan, or both.
However, maintaining the same image quality is not always reasonable. For example, if there
is more attenuating tissue at the end of the scan range, AEC tends to increase the tube current
on that area, even if a noisier image would be adequate for a diagnosis, such as evaluating lung
tissue at the liver level. Setting the maximum tube current value for modulation can reduce
the patient’s radiation exposure without compromising image quality.
With real time tube current modulation the possible contrast media in the scanned tissues
increases the current and radiation exposure of the patient. Also the use of bismuth shields on
the scanned area can mislead the AEC correspondingly.
AEC is recommended for paediatric body examinations; however it’s very important to know
how the AEC of the particular scanner in question works.
9
Figure 1. Steps of image reconstruction.
The image shows the steps of the CT image formation. The patient’s central axis direction
is called the z direction, and the x-y-plane is called the axial plane. SFOV (scan field of view)
indicates the x-ray beam size in the x-y-plane. DFOV defines the field of view of the displayed
images, and image matrix defines the pixel count in DFOV. The beam size in the direction
of the patient’s long axis, i.e. the nominal beam width, is the nominal size of the detector
elements used in scanning multiplied by their number. From the “raw data” collected during
scanning, displayed slices are created with different calculation processes at the workstation.
The thickness of the viewed slices, i.e. the reconstructed slice thickness, can be selected as
needed.
10
Pitch
Pitch is the shift of the scanner table during one rotation of the x-ray tube divided by the nominal
beam width. With older scanners the speed of the scanner table shift may be described by the
factor table feed (mm) per one rotation of the x-ray tube instead of pitch factor. Pitch can be
calculated by dividing the table feed with the nominal beam width.
When using AEC, the tube current is typically automatically reduced when pitch is reduced
and thus the patient dose and image noise does not change. Accordingly, the tube current will
increase automatically if the pitch value is raised. With AEC, the pitch value can usually only
affect the scan speed. If AEC is not used, one can make the scan faster and reduce the patient
dose by increasing pitch.
With AEC it is reasonable to use pitch 1. For larger children it is possible to use pitch value
higher than 1, to make the scan faster and potentially reduce movement artefacts.
If a low enough dose level cannot be achieved with AEC, dose level can be reduced by using
the lowest possible fixed mA, a short rotation time and high pitch. This approach might be
needed with small patients and when scanning extremities.
If the scanner automatically adjusts the current used while the pitch-factor is changing,
also the tube loading (mAs) used in scanning will be changed. Some scanners indicate the
tube loading using the effective mAs value (Eff.mAs) that is maintained constant against the
pitch value.
Tube loading [ mAs ]
Effective mAs = (1)
pitch
3.2 Dose and image quality
CT dose distribution in patient differs from the (projection-) x-ray imaging. Inner organs typically
receive a relatively higher dose in CT than in a conventional x-ray image. This is why one should
pay attention to the special features of CT scanning while optimizing the scan.
Technical image quality

Technical image quality is numerically indicated (and so quantifiable) and should be measured
as objectively as possible. It is most simply described with noise and contrast. The contrast-
to-noise ratio is obtained by dividing the difference of the CT numbers (e.g. test object’s CT
number compared to background in phantom) of the regions of interest (ROI) by the standard
deviation (SD) of the CT numbers. The SD of the CT numbers on a uniform ROI numerically
describes image noise. These quantitative image quality parameters are typically measured from
a specific test phantom or visually evaluated from details of the test phantom. The technical
features and scanning parameters of the scanner affect the technical image quality.
Technical image quality parameters can be used in evaluating the clinical image quality.
The clinical sufficiency of the image quality should always be evaluated by a radiologist.
Noise affects clinical image quality most when low contrast resolution is important e.g. for the
abdomen. In bone and lung areas, image noise is not as critical because these objects have high
contrast compared to surrounding tissues. Often bones and lung can be scanned with low dose.
11
Figure 2.
Example of a test phantom that can be used for
assessing low contrast test objects (lighter circles in the
image).
Other technical image quality parameters typically
assessed from CT images include accuracy of CT
numbers (HU-units), image uniformity and resolution of
small high contrast details. High noise level appears as
graininess in the image. Low dose level results in a high
noise level in the image, and makes it harder to see
image details behind the noise.
Image noise is inversely proportional to the square root of dose, i.e. the noise will be reduced
by a factor 0.7 if the dose level is doubled.
Dose quantities
The CT dose volume index CTDIvol describes the average radiation dose on the scanned area,
measured in a standard test phantom. This test phantom is an acrylic cylinder with the diameter
of 16 cm (head) or 32 cm (body). The weighted dose length product DLPw, is the product of
CTDIvol and the length of the scanned area d. In the dose displays of the newest scanners both
CTDIvol and DLPw are usually given. The accuracy of the dose display reading is verified by
regular measurements, as part of the quality assurance of the scanners.
DLPw
DLPw = CTDIvol · d or CTDIvol = (2)
d
12
Figure 3. Dose distribution with the same scan parameters in different size patients and standard
phantoms. (Image by: ImpactMC, CT Imaging GmbH, Erlangen, Germany)
Same scan parameters means that the dose display of the CT scanner would show the
same dose level (measured in 32 cm phantom) for both patients. In this image there is a
child in the upper left and an adult on the upper right corner, a 16 cm phantom (head) in the
lower left and a 32 cm phantom (body) in the lower right corner. The dose distributions are
colour coded so that blue means low dose level and yellow means high dose level. It can be
seen from the image that the dose level indicated by the dose display (i.e. dose measured in
32 cm phantom, lower right) doesn’t describe the real dose distribution in the patients very
well.
Since the CT scanner dose display indicates the radiation dose in a cylindrical standard size
acrylic phantom, it doesn’t take the patient’s size into consideration, and thus it doesn’t indicate
reliably the real radiation dose received by the patient. For this reason, determination of the
patient’s organ doses, effective dose and risk requires evaluation by a medical physicist.
With the same scanning parameters the dose level indicated for a 16 cm phantom is
approximately double compared to dose measured with a 32 cm phantom.
13
When a CT-examination is planned and the scanning parameters are changed, the effect on
the dose parameters can usually be seen on the dose display. The real time mA-modulation
can further change the dose values during the scan. The final dose parameter values can be
reviewed after the examination from the dose report that can be saved to the image archive.
To interpret dose parameters the information of the phantom size used for calculation (16 cm
head or 32 cm body) is needed. If the size of the dose phantom is not given, it should be verified
from the scanner guidebook or requested from the CT-device supplier.
It’s good to remember that the radiation dose of the examination cannot be reliably estimated
based only on the mAs values, since different CT scanners have different kinds of filters
depending on the scanner brand and model, and scanning is performed with different voltage
levels. For these reasons, there may be big differences in the radiation output (mGy/mAs).
Iterative reconstruction
Reconstruction of CT images is based on filtered back projection (FBP). During the last few years
iterative reconstruction has become more common in image reconstruction. In this method image
is reconstructed by inverse problem solving and the image noise is reduced using modelling
and repeating the calculation process. Reduced noise allows lowering the patient dose while
the image quality remains adequate for diagnosis. Images produced by iterative reconstruction
differ in appearance to some extent from CT images produced by FBP.
The iterative reconstruction algorithms and methods used by the scanner manufacturers
(e.g. AIDR, ASiR, iDose, IRIS, Safire, VEO) differ. The latest iterative methods are model based,
taking into account the physical features of the CT device and the scanning procedure. For this
reason, the image quality (especially noise) can be improved considerably. This is a new and fast
developing area of CT technology, and so there aren’t yet publications that cover it, especially
regarding the diagnostic advantages of the model based iterative methods.
The introduction of the iterative reconstruction program must be done by cooperation
between medical physicists and radiologists in order to maintain the image quality diagnostic
and to get the full advantage in dose optimization.
3.3 Using radiation shielding

The superficial tissues in the scan range, that are sensitive to radiation e.g. mammary glands,
thyroid gland and lenses of the eye, can be protected with bismuth shields. The aim is to filter
off the low energies of the beam that would otherwise absorb in to the patient’s tissues. Bismuth
shields may cause artefacts in superficial tissues under the shields, so using shields must be
considered case specifically. A 1 cm thick cotton wool layer between the shield and patient’s
skin considerably reduces the artifacts. On the other hand, a too thick layer will reduce the
shielding effect. Bismuth shields also affect the overall image quality, because they also partly
prevent the radiation that has already gone through the patient from reaching detectors and
so part of the patient exposure is lost and image quality is compromised. It’s important that the
use of bismuth shields is carefully considered, and they should always be chosen according to
the patient’s size so that they only cover the appropriate area. New CT scanners use different
techniques that can reduce the dose e.g. to breasts by decreasing the tube current when the
tube is in the anterior side of the patient.
14
Since AEC programs use attenuation
information of the planning image for
calculating the tube current beforehand,
the bismuth shields should be placed after
the planning image. This prevents the
unnecessary increase of the tube current
above the shield. If AEC changes the tube
current also in real time during scanning,
using a shield on the scan range may result
in an unpredictable dose level, so their
use is usually not recommended in these
circumstances.
Lead shields can be used outside the
scan range for protecting e.g. thyroid gland,
mammary gland for girls and testes for boys.
Shields should not be on the selected scanning
area on the planning image, since the AEC
program utilizes the density information
obtained from the planning image.
Bismuth shields affect to AEC acoording to the technique used in each scanner, and thus
the proper use of shields should be assured in operation training.
If the presence of the child’s parent or other companion in the scanning room during the
examination is necessary for a successful examination, the person must be carefully protected
using a lead apron, thyroid gland shield and, if possible, lead glasses. This person should be
placed in a point where scattered radiation is at minimum. Typically most of the radiation
scatters to the 45 degree angle between the patient and scanning aperture, and least beside
the aperture, where the CT scanner’s own structures stop part of the radiation.
3.4 Preparation for the examination

Careful preparation of the CT exam reduces the risk of not succeeding in making a useful image.
Procedures needed in preparation of a paediatric examination are listed below.
Reduce restlessness
• Describe the examination process to the child and the parent beforehand.
• One appropriately protected parent stays with the child, if needed.
• Sedation or general anaesthesia, if required.
Avoid pain
• Place the cannula in good time before the examination.
15
Practise cooperation
• Before the examination practising e.g. breath hold.
• Getting used to the table movements.
Prepare/plan the scanning carefully
• Defining carefully the scan range from the planning image.
• Patient positioning in the middle of the scanning aperture is very important for the correct
dose modulation function. Off-centering may increase the patient dose considerably and
worsen the image quality.
After the examination it is advisable to pay attention to the child, and it is usually a good idea
to give a small reward, e.g. a sticker or a coloured pencil.
16
3.5 Literature
1 American Association of Physicists in Medicine (AAPM). Use of bismuth shielding for the
purpose of dose reduction in CT scanning. AAPM Public Position Statement, PP 26-A;
2012. http://www.aapm.org/publicgeneral/BismuthShielding.pdf
2. Brenner DJ, Ellis CD, Hall EJ, Berdon WE. Estimated risks of radiation induced fatal
cancer for pediatric CT. American Journal of Roentgenology 2001; 176 (2): 289–296.
3. Hall P, Adami HO, Trichopoulos D et al. Effect of low doses of ionising radiation in infancy
on cognitive function in adulthood: Swedish population based cohort study. BMJ 2004;
328: 19.
4. Kalra MK, Dang P, Singh S, Saini S, Shepard JA. In-plane shielding for CT: Effect of off-
centering, automatic exposure control and shield-to-surface distance. Korean Journal of
Radiology 2009; 10 (2): 156–163.
5. Land CE, Tokunaga M, Koyama K, Soda M, Preston DL, Nishimori I, Tokuoka S. Incidence of
female breast cancer among atomic bomb survivors, Hiroshima and Nagasaki, 1950–1990.
Radiation Research 2003; 160 (6): 707–717.
6. Tack D, Gevenois PA (ed.) Radiation dose from adult and pediatric multidetector computed
tomography. Berlin: Springer-Verlag; 2007.
7. Yu L, Bruesewitz MR. Optimal tube potential for radiation dose reduction in pediatric CT:
principles, clinical implementations, and pitfalls. Radiographics 2011; 31: 853–848.
17
4. Vendor specific features
This section gives more scanner-specific information from scanner manufacturers about matters
related to optimization and paediatric examinations.
4.1 General Electric (GE)
Using radiation shielding

Shielding should be placed after planning image, when their effect on AEC can be avoided.

In the GE 3D modulation method, the tube current is controlled both while the rotation angle
of the tube is changing (xy-modulation) and parallel to patient’s z-axis (z-modulation). In
automatic exposure control the latest planning image is always used for calculating the required
mA values. If the latest planning image is a PA (or AP) projection, the dose level is a bit lower
than if a lateral planning image is used as the basis for modulation. For the proper function
of the mA modulation it is very important to center the patient in the middle of the scanner
aperture. Wrong centering increases the patient’s surface dose considerably and also affects
the image quality.
The modulation technique used by GE (AutomA and Smart-mA) uses noise index (NI) for
describing the noise level in an image filtered with a standard reconstruction filter in the
middle of the scanned object. The NI value determines the accepted noise level in images and
while selecting it the scan range, indication and the patient’s size must all be considered. The
tube current used for each rotation (mA-table) is defined with the NI value. For larger sized
patients, a higher tube current is used in order to reach the desired image quality. The lower
and upper limit for the used mA range is selected separately under the tube current menu (min
mA, max mA). Before scanning it should be verified from the mA table that modulation works
reasonably in the given mA range and is not saturated at the given max. or min. value. While
selecting the NI value it must be considered that it is always defined for a specific reconstructed
slice thickness. The software sets the mAs so that the user defined NI value of the images with
slice thickness of the first reconstruction series is achieved. If the reconstructed slice thickness
of the images used for diagnosis differs from the first reconstructed image series’ (that is used
for determining the image quality) slice thickness the desired NI must be altered accordingly.
Typically the diagnistic slices are thicker so the NI can be higher for the first reconstruction
series.
SFOV
The selection of SFOV defines the type of the beam shaping filter (small, medium and large).
The lowest possible SFOV setting should always be used in examinations, taking into account
the patient size.
GE has two different iterative reconstruction methods; ASiR (adaptive statistical iterative
18
reconstruction) and the newer VEO. With ASiR the percentage of the iterative weighting of the
reconstructed image can be chosen, e.g. 40% ASiR weighting uses 40% iterative reconstruction
and 60% conventional reconstruction technique (FBP). In the first ASiR version image noise is
reduced by selecting manually the ASiR percentage. Reduced noise allows lowering the dose by
increasing NI while using modulation, or by decreasing tube current while using fixed scanning
parameters. Conversion factors must be used in calculating the amount of the changes. In the
new ASiR version the desired dose saving percentage is set, compared to the images without ASiR
using DRG (dose reduction guidance). After DRG selection, the scanner calculates automatically
the needed ASiR percentage and tube current so that the noise level stays the same as in images
acquired without ASiR. With small children the appropriately optimized scanning dose level
is already low, and then the performance of DGR should be verified and ASIR percentage set
manually, if required. It is recommend to define the scanner specific practice together with a
medical physicist and an operations consultant from the scanner supplier. Selecting the dose
saving and ASiR percentage is always an indication a specific solution. VEO is a new model
based iterative reconstruction method and the potential dose saving is larger than with ASiR.
The calculation time for VEO is long at present, and this restricts using the method in urgent
examinations, for now.
4.2 Philips

If bismuth shields are used they should be placed on the scan area after the scout scan, so that
they won’t affect the AEC. D-DOM ACE changes the tube current during scanning while the
tube rotation angle is changing (x-y modulation), and while using this program bismuth shields
should not be used in any situation.
Classification of scanning protocols

In Philips scanners, protocols are classified into groups: Infant (under 18 months), Child and
Adult.
Dose display
In infant protocols’ body region the dose calculation is based on a 16 cm diameter phantom,
in other body area protocols on a 32 cm phantom, and in all head area protocols on a 16 cm
diameter phantom.
Scanning parameters
When small objects are scanned the matrix size is important. Therefore, especially with children
it is reasonable to use 768 x 768 and 1024 x 1024 image matrices. It is beneficial to change also
the resolution to High or Ultra High mode with larger matrix size. Typically in examinations
where a large matrix size can be profitable (e.g. lungs, inner ear, skeleton) mAs value does not
have to be high, because the object’s contrast is high with respect to the surrounding tissues.
Using High and Ultra High mode in paediatric examinations for other than high contrast
objects is not recommended, because it requires increasing the mA value in these examinations.
19
While using a larger matrix size for reconstruction a sharp-filter is used and the slice thickness
is kept low. The effect of changing the scanning parameters on patient dose can be tested before
examination by using the simulation feature.

DoseRight (prev. ACS, automatic tube current control) suggests a mAs value individually for
each patient that is user-editable even in the scanning phase. The scanner compares the water
equivalent diameter calculated from the planning image to the reference patient’s size of the
scanner. If the patient has a larger size than the reference patient, the scanner suggests a
higher mAs value and vice versa. The following table shows the reference patient sizes set to
the newest scanner versions:
Infant Child Adult
Boby 16 cm 20 cm 33 cm
If the patient size differs considerably from the reference patient, the scanner gives the user a
notice, and the protocol can be changed before the actual image series begins.
• Z-DOM adjusts the tube current in z direction. The current profile is calculated before
the examination on the basis of the planning image. The direction of the planning image
(lateral, PA or dual) does not affect the AEC, and a planning image of one direction is
sufficient. When using Z-DOM user sees the mAs min-mean-max-values before the scan
and can adjust the max. value (thus, determining the maximum limit for mAs). When using
Dual scout the proper tube current is calculated on the basis of the first image (lateral).
• D-DOM modulates the tube current while the tube rotation angle is changing (xy-
modulation). Adjustment is real time, i.e. the attenuation difference between x- and
y-direction is calculated based on the previous round.
• DoseRight Cardiac is a dose saving method used in cardiac helical scans. The phases to
be reconstructed are defined before scanning, and the tube current is decreased for the
areas that are not included in the phases to be reconstructed.
The following recommendations apply to Brilliance 64 version <3.5 and version iCT <3.2
scanners:
• infants and small children: Z-DOM (not DoseRight)
• pre-schoolers: Z-DOM (not DoseRight)
• older and larger sized children
– Z-DOM (not DoseRight): neck/lung region
– D-DOM or without modulation: upper stomach/pelvis and shoulder/neck
• adult-sized children
– DoseRight or DoseRight+Z-DOM: neck/lung region
– DoseRight + D-DOM: upper stomach/pelvis and shoulder/neck.
The following applies to Briliance 64 version ≥3.5, iCT version ≥3.2 and Ingenuity CT scanners:
• In addition to the above mentioned recommendations DoseRight function (automatic
exposure control) can be used for all children, since these scanner versions have references
also for small children.
20
The user can choose the iDose level in the scanning protocol: 1–7. The iDose level does not
directly indicate the achieved dose saving as a percentage, but the dose reduction is done by
adjusting mAs to match each iDose level according to the examination type. A table created for
optimizing protocols gives the mAs levels corresponding to iDose-level for each examination
type. It is recommended to use at least one degree sharper filter with iDose.
4.3 Siemens

The AEC program uses attenuation information of the planning image for calculating the
sufficient tube current beforehand. Bismuth latex shields should be placed only after planning
image is taken. Since the tube current is adjusted also during scanning, it is important that the
bismuth shields cover only the appropriate area and the shields should be chosen according
to the patient’s size.

In the Siemens CareDose 4D modulation method the tube current is adjusted both while the
rotation angle of the tube is changing and along the patient’s long axis (z-modulation). The
software calculates the attenuation characteristics of the patient from the planning image and
adjusts the tube current for each rotation before scanning. During scanning the thickness of
the patient is inspected automatically real time in different directions and the tube current is
adjusted even more precisely. For example, in lateral projection the patient can be thicker, and
the current is adjusted higher than in AP projection.
In the CareDose 4D method, the only user adjustable parameter is Quality Ref mAs. The
value of this parameter determines the noise level of the patient image. The image quality
set by the Quality Ref. mAs is the image quality achieved for a normal sized (75 kg) reference
patient when the given mAs-value is used. If the patient to be scanned is smaller than this, the
mAs value is adjusted lower than the defined Quality Ref mAs value. In the same way, for a
patient who is larger compared to a normal sized patient the mAs value is raised higher than
the defined quality Ref mAs value. In paediatric protocols a normal sized patient is 25 kg and
CareDose 4D works well with children up to 35 kg. If a child is remarkably larger (more than
40 kg), AEC can raise the tube current too high. Therefore, from the radiation protection point
of view it is more reasonable to use a scanning protocol designed separately from the adult
protocol for larger paediatric patients.
However, the newest scanners and some update versions of older scanners use an adult
patient weighing approximately 70 kg also as paediatric reference patient. This should be
considered while planning scanning protocols. The reference patient used in the software version
should be verified before planning the paediatric scanning protocols.
The minimum tube current for a 16-slice scanner is 28 mA and for a 64-slice scanner 20 mA.
CareDose 4D can adjust the tube current only higher than this.
Also a Care kV feature may be available in the new Siemens scanners. Based on the planning
image and the selected scan type, the most suitable scanning tube voltage and the appropriate
21
Quality Ref mAs value are indicated. When using the Care kV feature, it should be always
ensured that it works appropriately and that it does not (for example), raise the scanning
voltage higher than necessary.
New Siemens scanners have the option of the X-care feature with which the tube current
can be lowered while the tube is in front of the patient and thus dose to e.g. eyes, thyroid or
breasts can be decreased without separate shielding.
Dose display
Siemens scanners calculate dose parameters also for paediatric examinations using the
information based on adult dose phantoms (head 16 cm and body 32 cm).
When using the iterative reconstruction program a 1–2 grades sharper filter (Kernel) can be
used because an iteratively reconstructed image is typically softer than an image reconstructed
with the back projection method.
4.4 Toshiba

Bismuth shields are placed after the planning image so they don’t affect the mA-modulation
calculation.
Classification of scanning protocols

Scanning protocols are classified into five groups: Adult, Child, Trauma, Whole body and Chest/
Pelvis. Toshiba scanners have a specific examination tree structure (Child) where the paediatric
examinations are defined. The children’s examination tree structure appears automatically
with paediatric patients (age is user selectable; default is 12 years)
Dose display
Dose display takes the used SFOV size into account: When SFOV is M or lower, the dose display
is based on doses calculated with 16 cm diameter phantom. With higher SFOV values the dose
display is based on a 32 cm diameter phantom.

In the SureExposure 3D program, the automatic tube current modulation is done both in z- and
xy-directions, on the basis of anterior and lateral planning images. The desired image quality
(SD = Low Dose+, Low dose, Standard, High Quality or High Quality+) is selected. The accepted
noise level in images depends on the selected image quality and the region to be scanned. For
paediatric patients there are specific dose modulation curves that are not dependent on the
adults’ curves.
Since SureExposure is based on the selected noise level, the tube current level depends on
many different parameters e.g. the used primary reconstruction kernel. For example, a sharp
22
FC13-kernel results in a higher tube current than softer ones, FC12 or FC10.
If the SureExposure 3D AEC is not used, the noise level and minimum and maximum
values for tube current can be selected separately in paediatric examinations. The scanner
then functions between these values. It’s also possible to switch off the xy-modulation and use
only Z-direction modulation.
The Toshiba AIDR+ automatic iterative calculation also works for paediatric patients.
Miscellaneous
For paediatric patients there are specific FC-kernels (Toshiba’s patent) both for babies (infant)
and child patients (child).
Toshiba scanners have a simulation mode (Scan Simulator) which the effect of can be used
for evaluating the effect of scanning parameters beforehand. Also an example image appears
for which the noise level depends on the scanning parameters used.
23
5. Head region
5.1 Head
The primary method for paediatric brain examinations is magnetic resonance imaging (MRI).
CT is indicated in an acute situation and when MRI is contra-indicated or in rare special cases
(e.g. CT angiography in imaging of an acute vascular lesion). Ultrasound examination with
good technique through anterior fontanelle is usually adequate for babies e.g. in assessing
ventricular size and cerebral problems due to prematurity. However, it should be remembered
that ultrasound does not reliably exclude all types of traumatic lesions possibly requiring
neurosurgical intervention. Cerebral ischemia in children is examined with MRI, and CT
perfusion imaging used for adults should not be used for children because of the excessive
radiation dose.
Head trauma imaging is discussed within item 10.1.
Indications
• Suspected acute brain incident (e.g. intracranial haemorrhage).
• Imaging ventricular size in patients with ventriculo-peritoneal shunt if the fontanelle is
already closed and MRI is not available (see item 5.2).
• Surgical intervention planning in cases of premature closure of the cranial sutures (see
item 5.3).
Performing the examination

The examination can be performed either with axial slices or with helical scanning. If possible
the gantry should be tilted to avoid the eyes. Image quality should be good enough to differentiate
grey and white matter (both the cortical and deep gray matter, i.e. basal ganglia). I.v. contrast
medium is very rarely indicated; any additional information needed is best acquired using MRI.
5.2 Head scanning in the assessment of ventriclular size

As long as the fontanelle is open even a little, the ventricular size can be determined by
ultrasound. MRI is preferred when examining older children. If it is not possible to use MRI,
CT should be performed as a low dose examination. It is important to note that fully optimizing
the scan parameters will result in patient dose significantly lower than in routine head scan.
Indications
• Suspected ventricular dilatation in ventriculo-peritoneal shunt malfunction.

Since the image quality is needed only for assessing the ventricle size, quite noisy images
are sufficient for diagnosis. Thus remarkably lower scanning parameters can be used than in
routine head CT. Scanning 4–5 slices at the ventricle level is enough for diagnostics. In the
planning image one slice is placed at the level of the ventricular shunt tip in order to confirm
its localisation.
24
5.3 Scanning the cranial sutures
Diagnosis of premature closure of the cranial sutures is usually clinical (Duodecim 2007).
If necessary, ultrasound examination or plain x-ray can be used according to radiologist’s
instructions. Ultrasound (at least a 7.5 MHz probe) has been shown to be reliable especially
for examination of single sutures (e.g. plagiocephaly). CT is usually needed only for surgical
planning.

CT is performed according the instructions of the institution performing the operation in
order to ensure adequate 3D-reconstructions. The image quality needs to be adequate only to
delineate the bony structures.
5.4 Literature
1. Dunning J et al. Derivation of the children’s head injury algorithm for the prediction of
important clinical events decision rule for head injury in children. Archives of Disease in
Childhood 2006; 91: 885–891.
2. Kuppermann N et al. Identification of children at very low risk of clinically-important
brain injuries after head trauma: a prospective cohort study. Lancet 2009; 374 (9696):
1160–1170.
25
6. Ear, nose and throat region
ENT region is usually scanned with CT to examine the bony structures; so the natural contrast
of bone allows low scanning parameters. A low-dose technique should be used especially when
scanning paediatric facial bones and paranasal sinuses. Scanning parameters should be
optimized according to patient’s size. With good optimization the patient dose might be as low
as with a few conventional radiographic images of the same region.
6.1 Paranasal sinuses

In complicated or chronic sinusitis CT provides a great deal more information of the anatomy
than conventional radiography. Also cone beam CT can be used for surgical planning.
Indications
• Surgical planning (FESS-TT): surgical mapping before operation, low-dose-CT
• Complicated sinusitis (epidural abcess or orbital subperiosteal abcess): surgical mapping
before operation if needed, to allow surgery in navigator control: low dose CT
• Suspected tumour: imaging bony structures in addition to MRI, good quality images of
both bone and soft tissues are needed.

Thyroid shielding is preferred when possible. Bismuth shielding for eyes cannot be used if
the surgery is performed in computed navigation control. Computed navigation is based on
recognising the facial surface which will be distorted if shielding is used.
The examination is performed as a helical scanning in axial direction, 1mm slices are
reconstructed in axial, coronal and sagittal planes using both bone and soft tissue algorithms.
Scanning voltage of 80 kV is usually adequate for all patients to delineate the bony structures
of the ENT region. When performing low dose scans the effective mAs can be 17–25 mAs
according to the child’s size.
6.2 Ear region

CT examination of the inner ear is a special examination that is requested mainly for surgical
planning.
Indications
• Cochlear implantation surgery: both MRI and CT are performed.
• Further imaging of an inner ear anomaly in addition to MRI. Many of structural anomalies
of the inner ear, such as anomalous auditory ossicles, or auditory canal atresia can only
be visualized with CT.
• Cholesteathoma is diagnosed clinically, imaging is needed only in special cases; CT is used
mostly for surgical planning.
• Complicated infection (e.g. subperiosteal or epidural abcess, sinus thrombosis): MRI is the
first-line examination for children, but CT might be needed to examine bone destruction.
26
In spite of developed reconstruction techniques, careful positioning of the patient head is
necessary to keep the scan length short in order to minimize radiation dose. Head should be
positioned straight to scan the temporal bones within the shortest possible scan range, from the
tip of the mastoids to the upper edge of the tympanic cavity. 0.5 mm slices are reconstructed in
true axial (skull base plane) and coronal planes. Both bone and soft tissue algorithms are used
to examine inflammatory lesions and congenital anomalies.
6.3 Facial bones

Facial fractures are uncommon in children because of the craniofacial proportions and small
paranasal sinuses. Coventional radiography of the facial bones has been found unreliable with
paediatric patients, and thus CT is the first-line examination when facial fracture is clinically
suspected. Suspected nasal bone fracture should not be imaged at all.
Orbital fractures are usually in the orbital roof in children under 7 years of age, and in the
orbital floor in older children because of the development of the paranasal sinuses. Orbital
fractures are rarely treated surgically in children, but inferior rectus muscle entrapment
(diagnosis made clinically) requires operation.
Indications
• Suspected clinically significant (requiring surgical treatment) facial bone fracture: low-
dose examination.

The examination is performed as helical scan (low-dose examination is recommended in the
literature, e.g. Morales et. al.). Axial, coronal and sagittal reconstructions are done both with
bone and soft tissue algorithms. A 1–2 mm reconstructed slice thickness may be needed to
examine the facial bones.
6.4 Literature
1. Morales JL, Skowronski PP, Thaller SR. Management of pediatric maxillary fractures.
Journal of Craniofacial Surgery 2010; 21 (4): 1226–1233.
2. Mulkens TH, Broers C, Fieuws S, Termote JL, Bellnick P. Comparison of effective doses for
low-dose MDCT and radiographic examination of sinuses in children. American Journal
of Roentgenology 2005; 184 (5): 1611–1618.
27
7. Neck region
CT of the neck is indicated only in special cases. The primary modalitities used in the examination
of this area are ultrasound and MRI. Ultrasound may not demonstrate the deep inflammatory
lesions, and in children the poor visualization of fat lines makes the soft tissue differentiation
in CT difficult. If MRI is not possible, contrast medium enhanced CT is recommended.
Spine trauma imaging is discussed in item 10.2.
28
8. Chest region
Helical scanning is usually used when performing CT examinations of the chest. Exception is
HRCT examination, which is performed with single thin slices of the whole lung or the region
of interest.
The use of contrast medium should be considered case-by-case basis. The use of i.v. contrast
may help to delineate mediastinal structures especially in small children, but should be
considered individually. With older children, and especially in control studies, it is rarely needed.
Contrast medium is not useful in detecting metastases in lung parenchyma.
CT angiography is an alternative option to MRI examination especially when examining
neonates (because of their small vascular structures). Lowering the tube voltage (kV) increases
image contrast, and is recommended especially in angiography. Lowering kV may reduce the
patient dose, but the increased noise may require increasing the tube loading (mAs).
Spine trauma imaging is discussed in item 10.2.
8.1 Thorax
Chest CT is usually used to diagnose and follow-up metastatic disease, but MRI is often better
in diagnosing mediastinal tumors. This is especially true for posterior mediastinal tumors
where involvement of the spinal canal can be crucial information. For pulmonary abscesses,
especially in follow-up, the accuracy of MRI is often adequate. When assessing pleural fluid,
ultrasound is the modality of choice, and CT has not been shown to give more information in
the characterization of the fluid.
Imaging of the bony structures is discussed in items 10.3. (Trauma) and 11.5 (Spine).
Indications
• Complicated infection
– For lung abscess detection/follow-up MRI – examination must be considered.
– Ultrasound for pleural fluid detection and characterization.
• Detection and follow-up of tumors and metastases. MRI is often a better choice, especially
with posterior mediastinal and chest wall tumors.
• Examination of parencymal and mediastinal structural anomalies (MRI examination
should be considered when possible). Ultrasound is often the best modality to visualize
the thymus.
• CT angiography:
– small vascular structures (less than 4 mm)
– pulmonary embolism
– vascular structures and heart, when MRI is not possible due to a long sedation or
metal implants.
29
With modern multi-slice equipment and fast scanning, breath hold is not necessary with small
children. Sedation often provokes athelectasis which may be difficult to prevent. In such cases
examining the child in prone position may aerate the athelectatic posterior parts of the lung.
Scan range is based on the indication. Usually the whole lung region from apices to just
below the pleural angles is scanned, but over-extending the scan range should be avoided. The
good natural contrast of the lungs allows the use of both low tube voltage (kV) and loading
(mAs) to reduce the patient dose. Tube voltage of 80 kV is usually adequate for children under
10 kg and 100 kV for children under 60 kg.
The need of i.v.contrast medium depends on the indication; lung metastases can be visualized
without. However, contrast medium should be used at least in the diagnosis of complicated
pneumonia and tumors. It can also be used to outline the mediastinal structures, especially
with small children (thymus). Unionized iodinated contrast medium with usually 300 mg I /
ml is used when needed. Volume of contrast medium is 1.5–2 ml/kg under 20 kg and 1–1.5 ml/
kg over 20 kg up to 50 ml. Injection rate depends on the size of the child, the vessel caliber and
the indication of the scanning; usually 1–3 ml/s. An automatic injector pump should be used,
if possible. Scan delay should be set to 20–30 s (depending on the size of the child), in order to
visualize all vascular structures.
8.2 High-resolution CT
High-resolution Computed Tomography (HRCT) with thin interspaced slices is a valuable tool
in the diagnosis of diseases of the lung parenchyma and the airways. HRCT is more sensitive
than plain film chest x-ray in the detection of diffuse interstitial lung disease, and more accurate
in characterizing them. The spectrum of the diffuse interstitial lung diseases is different with
children than with adults. Both clinical and radiological presentations are less specific with
children, and even histopathological differential diagnosis is not always straightforward; thus
accurate diagnosis is challenging.
The aim of an HRCT examination is to assess the presence and extent of the lung pathology,
characterize the lesions, restrict differential diagnosis and demonstrate a good site for biopsy.
The secondary pulmonary lobule is a fundamental unit of lung structure, the smallest
structure that is surrounded by connective tissue septa and can be visualized in HRCT. It
consists of approximately ten pulmonary acini with vessels, and understanding its anatomy is
essential in interpreting the HRCT examination and classifying the findings.
Indications for pediatric HRCT

• Normal chest radiograph with severe/unexplained lung symptoms, e.g.
– unexplained fever in an immunocompromised patient
– unexplained dyspnea/oxidation failure
– discrepancy between the pulmonary function tests and the chest radiograph.
30
• Abnormal but non-specific chest radiograph
– Ill-defined nodules, opacities or suspected interstitial lung disease
– biopsy site selection.
• Detection of bronchiectasis.
• Detection of sequelae of infection in a child with persistent symptoms
– bronchiolitis obliterans (BO) or bronchiectasis.
• Cystic fibrosis: staging, response to therapy (MRI feasible in follow-up).
• BPD (bronchopulmonary dysplasia): evaluation of the severity of the disease in special
cases.
• Follow-up of diffuse interstitial lung disease or airway disease.
HRCT technique
Breath hold is necessary to achieve good quality images, and this is why small children usually
need general anaesthesia and scanning in breath control. Good cooperation with the anaesthesia
personnel is important. For a successful examination the interval between breath holds can be
increased e.g. to 15–20 seconds.
If general anaesthesia is not possible due to the child’s condition, the examination can be
performed during quiet breathing, but with less accuracy. For expiratory images the child can be
scanned in lateral decubitus position, when the non-dependent lung appears as if in inspiration
and the dependent lung as if in expiration.
The axial HRCT images (1–1.5 mm) are reconstructed on a high spatial resolution algorithm
(edge/bone) in inspiration and expiration. The acquisition time should be as short as possible
in order to avoid the motion artefacts. The suitable slice interval in primary examination is
usually 10 mm in inspiration. In follow-up examinations the slice interval can be extended to
15–20 mm. The 3–5 expiratory slices are usually obtained so that the highest is acquired above
the aortic arch and the lowest approximately 1 cm above the upper diaphragm. The purpose
of the expiratory scan is to show the air trapping, thus worse image quality is acceptable than
in inspiration scans. This should be considered when adjusting the scan parameters. It may
be necessary to acquire few additional scans in a prone position due to athelectasis of the
posterior lung.
Image windowing should be adjusted according to the viewing devices and conditions.
However, it is important to always use the same windowing, to allow image comparisons.
If volumetric acquisition is needed in addition to HRCT, adequate HRCT images can usually
be reconstructed with modern scanners from the helical scan raw data, and a separate scan
is not needed. Since with edge-enhanced algorithms the image noise is higher, an increase of
the tube current may be necessary to maintain adequate diagnostic quality. Contrast medium
is not used with HRCT.
31
Figures 4 and 5. Example images of HRCT slices. The upper is during inspiration and the lower during
expiration.
32
8.3 Chest CT-angiography examinations
Cardiac structures can be better examined with ultrasound than CT, but the visualization of
the vascular structures can be inadequate. In neonates especially, abnormal vessels can be so
small that the resolution of MRI is not adequate either. CT examination provides information
also about the bronchi and lung parenchyma, which can be useful especially with neonates.
Indications
• Small vascular structures (less than 4 mm).
• Pulmonary embolism.
• Heart, when MRI is not possible due to long sedation or metal implants.

Anaesthesia in needed for small children, but breath hold is usually not necessary. Scanning
without contrast medium is usually not useful in CTA. In cardiac imaging the heart may
sometimes be scanned twice, if the enhancement of different structures at different phases is
needed.
The cubital vein should be used for fast injection to get sufficient concentration of contrast
medium in the desired strucures. Automatic injection should be used whenever possible.
Scanning delay depends on the interest: e.g. the pulmonary arteries or veins, the aorta or all
vascular structures at the same time. The delay depends also on the child’s size and the injection
site. Children have fast circulation, and therefore assessing the injection delay is difficult.
Automatic measurement of the contrast medium concentration should be used if possible.
When scanning very small children, the scan delay needed by the protocol might be too long.
In these cases, the required delay to the beginning of the scan must be estimated. If the scan
is started right at the end of the injection, all vascular structures are usually moderately well
enhanced. CTA is performed with a fast contrast medium injection and then flushing with a
fast saline injection. When scanning babies the injection rate of 1–1.5 ml/s is usually enough,
and with older children 2–3 ml/s. Injection rate of 4–5 ml/s may sometimes be necessary but
usually only for children over 10 years of age.
8.4 Literature
1. Copley SJ, Padley SP. High-resolution CT on paediatric lung disease. European Radiology
2001; 11 (12): 2564–2575.
2. García-Peña P, Boixadera H, Barber I et al. Thoracic findings of systemic diseases at High-
resolution CT in children. Radiographics 2011; 31 (2): 465–482.
3. Klussmann M, Owens C. HRCT in pediatric diffuse interstitial lung disease – a review
for 2009. Pediatric Radiology 2009; 39 Suppl 3: 471–481.
4. Kuhn JP, Brody AS. High-Resolution CT of pediatric lung disease. Radiologic Clinics of
North America 2002; 40 (1): 89–110.
5. Webb WR. Thin-section CT of the secondary pulmonary lobule: Anatomy and the image
– The 2004 Fleischner lecture. Radiology 2006; 239 (2): 322–338.
33
9. Abdominal region
Ultrasound is the modality of choice in paediatric abdominal problems. MRI has widely replaced
CT in abdominal imaging in paediatrics, although long scanning times often require general
anaesthesia. When examining abdominal tumours, CT is seldom used, and also in cancer follow-
up a combination of chest CT and abdominal MRI is good practice.
The small amount of abdominal fat and thus poor natural contrast make the visualization
of different structures challenging, and thus i.v. contrast medium is usually necessary. There is
no clear consensus in the literature about the filling of the intestine with radiolucent or radio-
opaque contrast medium, and the need depends on the indication of the scan. In Finland the
diagnosis of appendicitis is usually based on clinical findings, and ultrasound can be used to aid
in differential diagnostics. When examining paediatric acute abdomen, CT is seldom needed.
The exception is urinary tract calculi, which are rare in children. Then CT is performed as a
complementary examination to ultrasound with very low dose and without i.v. contrast medium.
To image abdominal vascular structures CT is still an important method, although the
indication is often related to complicated transplantation surgery. Also the use of gadolinium
in MRI examination can be contraindicated for these patients, because of the related risk of
nephrogenic systemic fibrosis (NFS).
Indications
• Trauma with clinical signs of abdominal injury or an unstable patient.
• Complicated infection when ultrasound does not give enough information and MRI is not
available.
• Suspected urinary tract stones, if ultrasound does not give enough information (low-dose
examination without contrast medium).

With modern fast CT scanners breath hold is not obligatory but it is recommended. The need
for anaesthesia should be considered case-by-case considering the child’s developmental level.
In most cases the examination is performed with i.v. contrast medium, and thus the child
should have venous access.
Filling the bowel with water or positive oral contrast medium should be considered case-by-
case basis. However, the children coming to CT examination are usually so ill that they cannot
drink the contrast medium, and trauma patients are always scanned without p.o.contrast
medium. Radio-opaque orally administered contrast medium can be utilized mainly to diagnose
abdominal abscess. The smaller the child is, the less he/she has abdominal fat and the more
difficult it is to follow the bowel loops. Thus it is difficult to see whether the fluid seen in the
image is inside or outside the bowel. 10 ml of water soluble 300 mg I /ml contrast medium
added to 400 ml water or juice can be used as oral contrast. This is given 20 ml/kg divided to
2 doses. The passage time of the contrast in small bowel varies a lot in children, but has been
shown to be approximately 1.5 hours. Thus, the first dose is given 1.5 hours and the second
15–30 mins before scanning.
Unionized i.v contrast medium of 300 mg I /ml is usually used, volume of 2 ml/kg. Scan
delay depends on the child’s size, the cannula used and the injection rate. Different methods
34
to calculate the injection rate (ml/s) have been published, e.g.: weight (as kilograms) · 0.1 or
(injection volume (ml)) / (delay(s)-15 s).
Enhancement of the liver parenchyma is sufficient, when contrast is seen in the hepatic veins.
According to some studies, it takes 50 s from the beginning of the injection for small children
and 60–70 s for older children. If automatic measurement of the contrast medium concentration
is used, the ROI is placed on the upper part of the abdominal aorta, and scanning is started –
depending on the child’s size 30–50 s after the concentration in the aorta is sufficient.
Scan region is from diaphragm to symphysis, but depends on the indication. Scanning the
urinary tract is started from the upper end of the kidneys. Over-extending the range should
be avoided.
Reconstructed slice thickness of 2–3 mm is often good; even 4–5 mm (especially for bigger
children) is possible. Image reconstruction is usually soft tissues weighted. However, in trauma
cases bone reconstructions are also needed. Lung windowing may aid in showing air outside the
bowel. Reconstructions should be done also in the coronal plane, which may help in assessing
the bowel loops.
With children the small amount of abdominal fat makes assessing the bowel and its walls
more difficult. Sometimes it is impossible to differentiate the appendix from the intestine.
Differentiating a slightly inflamed intestinal wall from normal may also be tricky.
9.1 CT-angiography of the abdominal region
Indications
• Scanning abdominal vascular anomalies in special cases, if ultrasound does not provide
enough information and MRI is not possible or its resolution is not sufficient.

A large cubital vein should be used for injection to get sufficient concentration of the contrast
medium. Usually 2 ml/kg of 300 mg I /ml unionised contrast medium is used. Scan delay depends
on the child’s size, the cannula used and the injection rate. Automatic injection should be used
whenever possible. CTA is performed with a fast contrast injection and then flushing the
contrast medium with equally fast saline injection When examining babies, the injection rate
of 1–1.5 ml/s is usually enough, and with older children 2–3 ml/s. An injection rate of 4–5 ml/s
may be sometimes needed but usually only when imaging teenagers. It may be difficult to
determine the contrast delay, because of the fast circulation. Automatic measurement of the
contrast concentration should be used, if possible. The ROI is usually placed on the upper part
of the abdominal aorta. When scanning very small children the scanner’s shortest delay from
the aortic enhancement to actual scanning can be too long. Therefore the scan can be started
right at the end of the contrast medium injection, and all vessels are usually moderately well
enhanced. Per oral contrast medium is not used in CTA.
9.2 Literature
1. Frush DP. Pediatric abdominal CT angiography. Pediatric Radiology 2008; 38 Suppl 2:
259–266.
35
10. Trauma
CT has a central role in the imaging of severely injured children. The primary purpose of
imaging is to recognize life threatening injuries. However, paediatric patients should not be
scanned routinely with CT except in cases of high-energy trauma. The need of CT should always
be considered case-by-case basis. Instead of whole body scanning the paediatric trauma CT is
often focused to for example abdominal area.
Although the sensitivity of ultrasound is only about 55% also in children, combining it to
clinical examination can raise the sensitivity to 100%. However, ultrasound has a remarkably
larger role in paediatric imaging than in adults. When examining haemodynamically stable
patients and FAST US (focused assessment with sonography for trauma) is found normal, a
more thorough SLOW (second look if otherwise well) ultrasound examination has been suggested
instead of CT.
The decision of CT examination is done based on clinical findings (pain, haemodynamic
stability, neurologic signs). Also the related injuries affect the decision, e.g. spinal or pelvic
fracture may require CT.
10.1 Head trauma

Since especially in small children assessing the level of consciousness due to injury using
the GCS (Glasgow Coma Scale) scale is very unreliable, different rules have been created
to recognise those patients who need CT. The best validated guideline is the PECARN rule
(Pediatric Emergency Care Applied Research Network) which, however, recommends CT
examination quite easily. The following list of indications for head CT examination follows the
recommendations from a British study (Dunning 2009, CHALICE rule). Using these criteria
the sensitivity to predict a clinically significant brain injury was 98% and specificity up to 87%.
• GCS <14, or <15 if under 2-years old
• drowsiness or abnormal behaviour
• positive focal neurology
• witnessed loss of consciousness over 5 min
• history of amnesia over 10 min
• signs of basal skull fracture
• large (>5 cm) bump/bruise in scalp if under 2-years
• seizure after head injury
• recurrent/increasing vomiting
• increasing headache
• high-energy injury (e.g. traffic accident >50 km/h, falling >3 m)
• multiple injury.
10.2 Spinal trauma

When examining spinal trauma it should be remembered that children may have a significant
spinal injury without abnormalities in bony structures (SCIWORA).
36
Indications
Cervical spine
• Clinical suspicion of cervical spine fracture.
• Plain radiography equivocal or suspicion of unstable fracture.
In a multi-centre study (Leonard et al 2010) the recommended criteria which indicate increased
risk of cervical fracture were:
• loss of consciousness in trauma patient (also intoxication)
• focal neurologic findings
• complain of neck pain
• Torticollis
• substantial injury of the body
• predisposing condition (e.g. Down’s syndrome, spondyloepiphyseal dysplasia (SED))
• diving as trauma mechanism
• high-energy injury.
Thoracic and lumbar spine

• Plain radiography equivocal or suspicion of unstable fracture.
10.3 Trauma of thorax and abdomen
Indications
• Clinically suspected thoracic injury or laceration of parenchymal abdominal organ.
• Further evaluation of a parenchymal organ injury detected in ultrasound examination.
• Further imaging of findings in chest radiography.
10.4 Performing trauma examination

Head is scanned without contrast medium with helical or axial scanning. In helical scanning
the dose to the orbital lens may increase if overranging is not prevented. However, helical
scanning allows scanning of the facial structures at the same time if needed, thus the dose can
be lower than when scanned separately.
Cervical spine can be scanned with the collar in place. Scanning of the body should be
performed hands above head, if possible. Radiation exposure is lower if the body region is
scanned as one stack. The examination of the body is performed with i.v. contrast (2 ml/kg,
300 mg I /ml). Bowel contrast is not used. If CT-angiography (e.g. lower extremities) is needed,
is should be performed before the scanning of the body. Then the contrast for angiography can
be utilized in the imaging of the body region.
The thoracic scanning should be timed so that there is a good contrast concentration in the
major vessels.
Good enhancement of the parenchymal organs is essential in abdominal scanning. Scan delay
37
varies significantly depending on the child’s size and haemodynamics (ca 10–30 s for thoracic
and 50–70 s for abdominal region). Alternatively a two phase injection (2 ml/kg) can be used.
First 80% of the contrast is injected to enhance the abdominal organs and then 20% just before
the scan to enhance also the major vessels at the same time. The whole body is scanned as one
stack. Thus the abdominal organs are enhanced and there is also good contrast in the major
vessels. If renal collection system injury related to kidney trauma is suspected, delayed scan is
performed at kidney level after 5–15 minutes to see the possible extravasation of the contrast.
When examining babies, the injection rate of 1–1.5 ml/s is usually enough, and in older children
2–3 ml/s. Injection rate of 4–5 ml/s may be sometimes be needed but usually only in teenagers.
10.5 Literature
1. Dunning J et al. Derivation of the children’s head injury algorithm for the prediction of
important clinical events decision rule for head injury in children. Archives of Disease in
Childhood 2006; 91: 885–891.
2. Kuppermann N et al. Identification of children at very low risk of clinically- important
brain injuries after head trauma: a prospective cohort study. Lancet 2009; 374 (9696):
1160–1170.
3. Leonard JC et al. Factors associated with cervical spine injury in children after blunt
trauma. Annals of Emergency Medicine 2011; 58 (2): 145–155.
4. Maguire JL, Boutis K, Uleryk EM et al. Should a head-injured child receive a head CT
scan? A systematic review of clinical prediction rules. Pediatrics 2009; 124: e145–154.
5. Scaife E, Rollins MD. Managing radiation risk in the evaluation of the pediatric trauma
patient. Seminars in Pediatric Surgery 2010; 19: 252–256.
38
11. Orthopaedic examinations
Orthopaedic CT examinations are mainly performed for the surgical intervention planning.
The natural contrast of bones usually allows low tube current (80 kV) also for bigger patients.
On the other hand, it may be necessary to increase the scanning parameters if there is metal
inside the scan range from previous operations.
11.1 Measuring leg length discrepancy from the planning image

Leg length difference is examined clinically, and imaging is used when surgical intervention
(epiphyseodesis) is planned. The purpose of imaging is to find out whether the length difference
is more pronounced in the femoral or tibial region. A CT planning image is the recommended
imaging method due to its low radiation exposure. But because the scan is done supine, the
possible effect of the foot anatomy or mal-alignment of joints to the length discrepancy cannot
necessarily be seen.
Indications
• Evaluation of leg length discrepancy, when epiphyseodesis is planned, and the clinically
determined difference is more than 2 cm.

Measuring is done on the planning image. If possible, the planning image is acquired in the PA
direction. Often 80 kV is enough when bony structures are evaluated. The patient is positioned
with legs side by side in a similar way, pelvis as straight as possible, feet forward, no external
rotation, toes together (supports can be used around the knees or beside the feet).
11.2 Joints
It is difficult to place only one joint (e.g. ankle) in the isocentre of the scanner so that AEC
would work properly, and thus fixed scanning parameters instead of AEC should be considered.
Indications
• Further evaluation of fracture of the articular surface to plan surgical treatment.

Bony structures have high natural contrast, and thus lower scanning parameters can be used.
80 kV is sufficient as tube voltage, and a low mAs value is enough especially for scanning
small joints. The scan range should cover only the joint area with the whole fracture area and
associated growth plates. It may be necessary to scan the elbow beside the body, if it cannot
be raised over the head. Then both kV and mAs must be increased almost to the level used
for body scanning. The so called “extremity CT scanner” can be used, if available. It allows e.g.
scanning only the elbow without radiation exposure to the body.
39
11.3 Growth plate scanning for evaluation of growth arrest
Indications
• Evaluating the size of growth plate fusion detected in radiography, for surgical treatment
planning.

The scan range should cover only the growth plates of the joint of interest. It may be necessary
to scan the elbow beside the body, if it cannot be raised over the head. Then both kV and mAs
must be increased almost to the level used for body scanning. When scanning the extremities a
so called extremity CT scanner can be used, if available. It allows e.g. scanning only the elbow
without radiation exposure to the body. Slices parallel to the growth plate are reconstructed from
the scan. The percentage of the prematurely ossified part of the growth plate area is measured.
11.4 DDH in cast
Indications
• Verifying the position of the femoral head in a child with pelvic cast for DDH, when MRI
is not possible.

The cast stabilizes the child’s pelvis so that the examination can be performed without sedation.
The scan range is from the upper edge of the acetabulum just below the femoral head. The
scanning can be performed with 2–3 axial slices or a short helical scan. Low voltage (80 kV)
and low mAs are enough to see the bony nucleus in the bony acetabulum. If examination of the
cartilaginous structures is needed, it is done with a MRI examination.
11.5 Spine
Indications
• Equivocal or unclear fracture in plain radiography.
• Suspected unstable spinal fracture.
• Precise anatomical clarification for surgical planning for scoliosis or vertebral anomalies.

With fractures, the scan range is determined by the conventional radiography findings. Over
extending of the scan range should be avoided. For the planning of surgical treatment the
operating surgeon defines the scan range according to the extent of the instrumentation. Also
pedicles of normal vertebrae should be inside the scan range, and vertebrae should be able to
be named. If CT scan data is used to make 3D plastic models for surgical planning, raw data
should be saved and sent to the unit that makes the models.
40
12. Hybrid nuclear medicine examinations
Computer tomography can be utilized in hybrid nuclear medicine imaging for three different
purposes:
• attenuation correction
• anatomical localisation of the radio-medicine uptake
• as a diagnostic examination.
These differ from each other regarding the required image quality and thus the patient dose.
The required image quality must be defined beforehand in order to optimize the examination.
The predefined paediatric protocols of the device manufacturer are typically designed to produce
a good image quality, and they are not optimized in relation to the patient dose.
An attenuation correction map of the scan region can be produced with the CT equipment.
It allows correction of the radio-medicine’s radiation attenuation in the patient. The electron
density information of the CT image is used to create the attenuation correction map. Scanning
with very low current and dose is sufficient to produce the attenuation map. Practically, the
smallest scanning parameters that the CT scanner can produce can usually be chosen, especially
with small children. Radiation attenuation in small children is minimal and for this reason
in SPECT (single-photon emission tomography) examinations the attenuation correction is
usually not necessary. Computational attenuation correction methods can be used for fairly
homogenous objects (e.g. a child’s head). Attenuation correction is always required to analyze
PET (positron emission tomography) images.
To localize the pathological radio-medicine uptake, image quality should be good enough to
distinguish the anatomical structures. This image quality is often sufficient also to replace the
diagnostic control CT examination, especially in chest region.
Diagnostic image quality in nuclear medicine CT-examination is required only if it can
replace the diagnostic CT examination, that would be otherwise required. For diagnostic
purposes a noisier image quality is usually adequate in paediatric CT examinations than in
adults’ examinations. The required image quality should be optimized together with a radiologist
specialized in paediatric imaging. When defining the image quality, it should also be considered
whether a CT examination is needed e.g. for external radiotherapy dose calculation or future
image fusions. Using contrast with attenuation correction scanning may cause over-correction.
Therefore, i.v. contrast cannot usually be used with attenuation correction scanning.
Indications
• Most children’s cancer types accumulate F-18-fluoro-deoxy-glucose (F-18-FDG) which
is the most common marker used in PET-CT. In primary diagnostics PET-CT should be
considered with
– Hodgkin’s lymphoma
– non-Hodgkin’s lymphoma with a atypical primary tumour/metastases
– extramedullary leucaemia, if primary tumour is not known
– soft tissue sarcoma
41
– MIBG negative neuroblastoma
– yolk sac tumour
– Langerhans cell histocytosis.
• Follow-up imaging is indicated only if recurrence or metastazing affects the patient’s
treatment. With recurrent disease PET-CT can be used to plan radiotherapy or determine
the biopsy site.
• For treatment response PET-CT should be considered in non-Hodgkin’s lymphoma, when
a poor response is suspected in conventional imaging. Information about the activity of
the remaining tumour is then achieved.
• Combining CT to paediatric SPECT examination is rare, but in some cases, e.g. gamma
scanning of the stomach mucosa, CT localization can give more clarity.

The need for diagnostic CT should be found out before the hybrid examination to avoid
unnecessary imagning. It can often be replaced with the CT examination performed with the
nuclear medicine examination.
The nuclear examination should be performed first, if possible, because in most cases it
can be repeated if required. If the CT examination has been performed first, and the patient
moves during the nuclear medicine examination, the CT cannot be used. However, in PET-CT
examinations the CT is performed first, and PET scanning is planned based on it. Thus, if the
patient moves during the PET scanning, the previously acquired planning image and possibly
also the low-dose CT may be spoiled. Some of the isotopes used in PET examinations are fast-
decaying, and one cannot repeat the examination many times. Supporting the patient carefully
during sedation is also important for a successful examination.
The long scan times should be taken into account in nuclear medicine examinations. This
is why small children are typically scanned under general anaesthesia. The patient should be
carefully supported also for the scanning, since nerve damage has been reported after a long
anaesthesia.
The parents can stay beside the child during the nuclear medicine examination, and a suitable
entertainment (a DVD, a book…) is usually enough to keep the child calm. The parents don’t
need to leave the child alone for radiation protection reasons, but during the CT examination
they should move further away, if possible, and be shielded appropriately (a lead vest, thyroid
shield, possibly lead glasses).
12.1 Literature
1. Alessio AM, Kinahan PE et al. Weight-based, low-dose pediatric whole-body PET/CT
protocols. Journal of Nuclear Medicine 2009; 50 (10): 1570–1577.
2. Antoch G, Feudenberg LS, Beyer T, Bockisch A, Debatin JF. To Enhance or Not to Enhance?
18F-FDG and CT Contrast Agents in Dual-Modality 18F-FDG PET/CT. Journal of Nuclear
Medicine 2004; 45 Suppl 1: 56S–65S.
3. Barrington SF, Begent J, Lynch T, et al. Guidelines for the use of PET-CT in children.
Nuclear Medicine Communications 2008; 29 (5): 418–424.
42
4. Chawla SC et al. Estimated cumulative radiation dose from PET/CT in children with
malignancies: a 5-year retrospective review. Pediatric Radiology 2010; 40 (5):681–686.
Epub 2009 Dec 5.
5. Dacian V. Bonta, Richard L. Wahl. Overcorrection of iodinated contrast attenuation in
SPECT-CT: Phantom studies. Medical Physics 2010; 37: 4897–4901.
6. Fahey FH, Palmer MR, Strauss KJ, Zimmerman RE, Badawi RD, Treves ST. Dosimetry and
adequacy of CT-based attenuation correction for pediatric PET: phantom study. Radiology
2007; 243 (1): 96–104.
43
Cover image: Sampsa Kaijaluoto, layout: Marja Niskanen. September 2012.
Laippatie 4, 00880 Helsinki

Puh. (09) 759 881, fax (09) 759 88 500
www.stuk.fi
ISBN 978-952-478-791-8
ISSN 1799-9510
Helsinki 2013

Advice From STUK September 2012 Paediatric CT

Uploaded by

Copyright:

Available Formats

Advice From STUK September 2012 Paediatric CT

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advice From STUK September 2012 Paediatric CT

Uploaded by

Copyright:

Available Formats

ADVICE FROM STUK / SEPTEMBER 2012

Guidelines for paediatric

3.1 Scanning parameters

Scanning voltage and tube current

Beam width and slice thickness

Nominal slice thickness = slice thickness used in scanning

Technical image quality

3.3 Using radiation shielding

3.4 Preparation for the examination

4.1 General Electric (GE)

Using radiation shielding

Automatic exposure control (AEC)

Using radiation shielding

Classification of scanning protocols

Automatic exposure control (AEC)

Using radiation shielding

Automatic exposure control (AEC)

Using radiation shielding

Classification of scanning protocols

Automatic exposure control (AEC)

Performing the examination

5.2 Head scanning in the assessment of ventriclular size

Performing the examination

Performing the examination

6.1 Paranasal sinuses

Performing the examination

6.2 Ear region

6.3 Facial bones

Performing the examination

Indications for pediatric HRCT

Performing the examination

Performing the examination

9.1 CT-angiography of the abdominal region

Performing the examination

10.1 Head trauma

10.2 Spinal trauma

Thoracic and lumbar spine

10.3 Trauma of thorax and abdomen

10.4 Performing trauma examination

11.1 Measuring leg length discrepancy from the planning image

Performing the examination

Performing the examination

Performing the examination

11.4 DDH in cast

Performing the examination

Performing the examination

Performing the examination

Laippatie 4, 00880 Helsinki

You might also like