Radiation Protection Dosimetry (2020), pp. 1–8 doi:10.

1093/rpd/ncaa002

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

SKIN DOSE MAPPING IN INTERVENTIONAL CARDIOLOGY: A
PRACTICAL SOLUTION
M. Krajinović1,2, *, M. Dobrić3 and O. Ciraj-Bjelac1,2
1
School of Electrical Engineering, University of Belgrade, Studentski trg 1, 11000 Belgrade, Serbia
2
Vinča Institute of Nuclear Sciences, University of Belgrade, Studentski trg 1, 11000 Belgrade, Serbia
3
Clinic for Cardiology, Clinical Center of Serbia, Belgrade 11000, Serbia

*Corresponding author: [email protected]

Received 11 July 2019; revised 25 November 2019; editorial decision 2 January 2020; accepted 6 January 2020

Numerous cases of radiation-induced tissue reactions following interventional cardiology (IC) procedures have been reported,
resulting in the need for an optimized and personalized dosimetry. At present, there are many fluoroscopy units without Digital
Imaging and Communications in Medicine (DICOM) Radiation Dose Structured Report globally installed. Many of these have
not been updated yet, and may never be, therefore, the main objectives of this paper are to develop an offline skin dose mapping
application, which uses DICOM headers for the peak skin dose (PSD) assessment and to compare the PSD assessment results
to XR-RV3 Gafchromic film for common IC procedures. The mean deviation between the measured and the calculated PSD
was 8.7 ± 26.3%. Simulated skin dose map showed good matching with XR-RV3 Gafchromic film. The skin dose mapping
application presented in this paper is an elegant solution and a suitable alternative to XR-RV3 Gafchromic film.

INTRODUCTION DICOM header associating them with the imaging

study. Even though these reports have a lot of
Numerous cases of radiation-induced tissue reactions
information, they do not contain all the necessary
following interventional cardiology (IC) procedures
information required for an accurate assessment of
have been reported, resulting in the need for an
the PSD or for generating a skin dose map. Another
optimized and personalized dosimetry(1) . Direct dose
major drawback is the necessity of using optical
measurement by placing a dosemeter on the patient’s
character recognition for extraction of necessary
skin is a suitable method for determination of the
information.
peak skin dose (PSD), which represents the highest
DICOM image has a header that contains most of
dose at any portion of a patient’s skin during the
the data (DICOM attributes) for assessing the PSD.
procedure(2) . However, given that such measurements
Although DICOM standard defines all the necessary
are complex and time consuming, it was necessary to
attributes necessary for dose assessment, it does not
design and develop a simpler and more practical way
require that vendors use all attributes in DICOM
of assessing the PSD. A big step toward a better way
header, only the mandatory ones(5) . In 2005 Radi-
of assessing the PSD was made by the International
ation Dose Structured Report (RDSR) was added
Electrotechnical Commission (IEC) by introducing
to DICOM standard with the intention to provide a
the concept of the reference air kerma (Ka,r ), defined
standardized format of recording all the information
as the air kerma accumulated at the interventional
reference point (IRP), which for C-arm fluoroscopic related to the exposure parameters used for each
systems is the point along the central ray of the X-ray irradiation event undergone by the patient. RDSR
beam, 15 cm back from the isocenter toward the focal contains all the necessary technical, geometric and
spot(3) . Since X-ray beam is directed at different areas dosimetric data required to assess the patient’s dose
of the patient’s skin during the procedure, estimates as opposed to the DICOM header(6) .
of the likelihood of radiation-induced skin injury that Most vendors have developed vendor-specific
are based on Ka,r tend to overstate this risk(4) . online solutions for skin dose calculations. Care-
Utilizing Digital Imaging and Communications in Monitor by Siemens and DoseWise by Philips are
Medicine (DICOM) standard for the PSD assessment basic solutions only providing accumulated peak air
and calculating the patient-specific skin dose map kerma according to the current projection(7) , whereas
has advantage because the solution created for one Dose Map from GE(8, 9) and Dose Tracking System
interventional X-ray system can be implemented for from Canon (Toshiba)(10–12) are advanced 2D and
interventional X-ray systems from different manufac- 3D solutions, respectively. 2D and 3D skin dose
turers with slight modifications(5) . The proprietary mapping solutions which utilize DICOM headers
dose reports are vendor-specific images that have and/or RDSRs can be found in the literature(13–18) .

© The Author(s) 2020. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
 M. KRAJINOVIĆ ET AL.

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

At present, there are many fluoroscopy units Table 1. Public and private DICOM attributes used for skin
without RDSR globally installed. Many of them dose mapping.
have not been updated yet, and may never be(19) ,
therefore the main objectives of this paper are to Tag Attribute name
develop an offline skin dose mapping application
which uses DICOM headers for PSD assessment and Public
to compare the results of PSD assessment to XR-RV3 (0018,0060) KVP
Gafchromic film (International Specialty Products, (0018,1110) Distance Source to Detector
Wayne, USA) for common IC procedures. (0018,1164) Imager Pixel Spacing
(0018,1510) Positioner Primary Angle
(0018,1511) Positioner Secondary Angle
(0018,1700) Collimator Shape
METHODS AND MATERIALS (0018,1702) Collimator Left Vertical Edge
Patient population (0018,1704) Collimator Right Vertical Edge
(0018,1706) Collimator Upper Horizontal Edge
The patient population included in the study were (0018,1708) Collimator Lower Horizontal Edge
those who received percutaneous coronary interven- Private
tion (PCI) procedures at the Heart Catheterization (0021,0017) Source to Isocenter
Department of the Clinical Center of Serbia. A (0021,1007) Skin Dose Accumulation
(0021,100a) Copper Filter
total of 10 PCI procedures were analyzed. For each
patient: age, gender, height and weight were recorded,
using the information from the medical history of
the patient. This study was approved by the Ethics
Board of the Clinical Center of Serbia, by Decision development. The application has a simple design
No 692/11. intended to ensure the ease of use. User only needs to
choose DICOM files generated after the completion
of cardiac procedure and to insert cumulative dose
Fluoroscopy unit (i.e. cumulative air kerma at the IRP) after which a
The procedures were performed on two identical color-coded map of skin dose over the back of the 2D
Siemens Axiom Artis dFA digital fluoroscopy patient is generated.
systems located in two operating rooms. The basic Skin dose mapping was based on the attributes
characteristics of the system are: 80 kW power (public and private) gathered from DICOM headers
generator, 50–125 kV voltage, 10–800 mA X-ray tube of images. All necessary data required for skin dose
current, flat panel detector and 4 mm Al inherent mapping are listed in Table 1. Total Ka,r after the
filtration with additional filtration from Cu (0– procedure is not listed in Table 1 and can be found
0.3 mm). The frame rate of 7.5 f/s was used for all in proprietary dose report or recorded manually after
cine acquisitions. procedure.

Calibration of kerma-area product (KAP) meter Geometry

Calibration of dose meters integrated in fluoroscopy Patients of different sizes were modeled as 2D silhou-
units, in dosimetric quantities of PKA and Ka,r was ettes in Photoshop with a resolution of 1 × 1 mm,
carried out in accordance with the standard proce- as shown in Figure 1. These models only show spa-
dure for X-ray tube(20) . Semiconductor dosimeter R- tial distribution of the skin dose, thus their selection
100 Barracuda (RTI Electronic, Molndal, Sweden), does not alter the PSD. Algorithm assumes head
placed on the patient table and pad (in order to first supine position of the patient with the center of
include the effect of the table and table pad in calibra- the back in central longitudinal position of the table
tion factor), was used as a reference standard for the and that there is no table movement in the lateral,
determination of PKA and Ka,r , whereas the field size longitudinal or vertical direction, due to the lack
was determined by XR-RV3 Gafchromic film. The of this information in DICOM header. The height
calibration factor acquired for Ka,r used for the PSD of the table is assumed to be 15 cm shifted below
assessment contributes to expanded uncertainty with the isocenter for each irradiation event. Coordinate
17% (k = 2). system is assumed with an origin at the isocenter of
the C-arm. Position of the X-ray tube focal spot is
calculated in Cartesian coordinates using the distance
Skin dose mapping application
from the focal spot to the isocenter, primary and
We developed a skin dose mapping application in secondary angles. Since the X-ray tube and detector
Python programming language ver. 3.7. Tkinter are always positioned opposite each other in a C-
library was used for graphical user interface (GUI) arm, central position of the detector is easily found

2
 SKIN DOSE MAPPING IN INTERVENTIONAL CARDIOLOGY

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

Figure 1. Patients figures as function of BMI (BMI values are in kg/m2 ).

using the position of the X-ray tube focal spot and the Application validation
distance from the source to detector. In order to find
XR-RV3 Gafchromic films calibration was per-
the affected patient area, it is necessary to find four
formed in the Secondary Standard Dosimetry
corners of detector. The affected skin location by X-
Laboratory of Vinča Institute of Nuclear Sciences
ray field is then found by finding intersection points
between the patient plane and the lines going through (VINČA). Calibration setup was done as described
X-ray tube focal spot and four corners of detector. in the paper of Farah et al.(23) , where comprehensive
Once four points in the patient plane are determined, characterization of XR-RV3 Gafchromic films was
a polygon can be defined, and thus the affected skin conducted. Calibration was carried out free-in-
area is found. air, where films were placed on a 0.5 cm-thick
PMMA support providing negligible backscatter and
irradiated at 1 m from the focal spot. Films from the
Skin dose calculation same batch were cut in small pieces of 2 x 2 cm and
After the affected skin location is identified, dose is aligned to the central flat region of the X-ray field.
calculated using Equation 1, where Ka,r is air kerma All film pieces were oriented so that the yellow side
at the IRP corresponding to fluoroscopic and cine of the film was in direction of the focal spot and dose
acquisition of each irradiation, CF is calibration fac- delivery was monitored using an ionization chamber.
tor for Ka,r determined annually from measurements The dose interval used to form the calibration curve
made during the annual quality control tests, dIRP is ranged from 0 to 5 Gy. Scanning was done after 24 h
the distance form the source to the Interventional of exposure. The Hewlett-Packard Scanjet G3110
Reference Point (60 cm in this case), dpatient is the was used for scanning the films. Color images were
distance from the source to each pixel of the affected acquired using 16 bits per channel (RGB images) with
area, BSF is backscatter factor interpolated from a resolution of 600 PPI in TIF format. Films were
the coefficients of Benmakhlouf et al.(21) , MAEC analyzed in the Python programming language using
is the ratio of mass energy absorption coefficients only the red channel of the scanned image, as this
for skin to air interpolated from the NIST Physical ensured maximum sensitivity. In order to calculate the
Reference Data Library (skin is modeled as ICRU PSD, a square region of interest (ROI) was formed
four component)(22) : with sides of 225 px (ROI was ∼1 cm2 ). The ROI
thus formed was shifted along the image, and then
the mean value of the red channel inside ROI was
 2 calculated, resulting in the mean pixel value (MPV)
dIRP
Dose = Ka,r × CF × × BSF × MAEC. of the red image component. Two unirradiated film
dpatient pieces were used in order to obtain the unexposed
(1) MPV. The calibration curve was obtained using
polynomial
 fitting based on reflectance, defined as
log MPVunexposed /MPVexposed , because it showed
Beam attenuation due to the presence of table
and table pad is included in CF. Since the fluoro- the best performance(23) .
scopic acquisitions are not recorded (only cine acqui- During cardiac procedures XR-RV3 Gafchromic
sitions are recorded), it is assumed that fluoroscopic films were placed below the patient in the region of
kerma is equally spread on each projection used for the upper part of the patient’s torso, with the longer
cine acquisition. Fluoroscopic kerma is calculated by side of the film along the cranial-caudal axis, and the
subtracting the total cine Ka,r from the total Ka,r , yellow side of the film oriented in direction of the
which can be found in a proprietary dose report or X-ray tube. PSD was obtained using the same Python
recorded manually after the procedure. routine used in film calibration. The ROI with the

3
 M. KRAJINOVIĆ ET AL.

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

Table 2. Patient population.

Patient Age (years) Height (cm) Weight (kg)

1 67 167 103
2 59 156 60
3 55 185 110
4 55 180 96
5 74 186 90
6 47 174 85
7 65 175 87
8 78 183 89
9 53 168 85
10 53 186 99
Mean ± Std (minimum–maximum) 60 ± 10 (47–78) 176 ± 10 (156–186) 90 ± 13 (60–110)

The last row represents the mean ± standard deviation (minimum–maximum)

lowest MPV value was used in order to find the PSD. • Entering cumulative dose, i.e. cumulative air
It should be pointed out that the film does not disturb kerma at the IRP (mandatory to enter in order
interventional procedure in any way, nor does it create to obtain skin dose map).
additional discomfort to the patient. • Displaying skin dose map.
The application generates a skin dose map fast
Geometrical comparison between XR-RV3 (e.g. it takes a few seconds to generate a skin dose map
Gafchromic film and simulation from 25 DICOM files). The skin dose map is gener-
ated in the right side of GUI every time the ‘SKIN
During the coding process geometrical verification DOSE MAP’ button is pressed. A colorbar is shown
was done in order to validate simulated skin dose right to the patient skin dose map for fast visual
map. XR-RV3 Gafchromic films were scanned as inspection of skin dose distribution. The calculated
8 bits per channel color images and were converted PSD is shown in the left corner of the skin dose map.
to grayscale images. Exposure, contrast and levels
were modified in order to enhance visual difference
between the exposed and the unexposed part of the XR-RV3 Gafchromic films calibration
scanned images. The measurement scale is shown on XR-RV3 Gafchromic calibration curve for determin-
all images for better comparison. ing skin dose is shown in Figure 3. Fitting equation is
as follows:
RESULTS
Patient population and fluoroscopy Ka,r Fit = p1 × Ref 3 + p2 × Ref 2 + p3 × Ref + p4 , (2)
Overall, ten PCI procedures (eight men and two
women) were analyzed in this paper. The mean age, where p1 , p2 , p3 , p4 are coefficients of a third-degree
height and weight of patients are shown in Table 2. polynomial and Ref stands for reflectance. Coeffi-
Average contribution of fluoroscopic Ka,r to total cients have the following values: p1 = 5.234 × 105 , p2
Ka,r was 57.2%, whereas average contribution of cine = -7.841 × 104 , p3 = 1.358 × 104 , p4 = -24.88.
Ka,r to total Ka,r was 42.8%. The mean total Ka,r , mean
fluoroscopic Ka,r , mean cine Ka,r and total PKA of
procedures are shown in Table 3. Geometrical comparison between XR-RV3
Gafchromic film and simulation
Skin dose mapping application Geometrical comparison between XR-RV3 Gafchro-
mic films and simulation for two different procedures
The GUI of developed application is shown in
is shown in Figure 4. Clearly there is resemblance
Figure 2. This application enables:
which proves that geometrical solution is accurate. It
• Choosing DICOM files to be used for generating should be noted that the scanner used for geometrical
skin dose map. comparison was not used for calibration purposes
• Entering patient height and weight for patient- due to the presence of horizontal lines on scanned
specific skin dose mapping. images.

4
 SKIN DOSE MAPPING IN INTERVENTIONAL CARDIOLOGY

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

Table 3. Total, fluoroscopic, cine K a,r and total PKA for all procedures.

Patient Total Ka,r (mGy) Fluoroscopic Ka,r Cine Ka,r (mGy) Total PKA
(mGy) (μGym2 )

1 447 120 327 3388

2 364 225 139 2874
3 871 494 377 7432
4 476 226 250 3106
5 901 626 275 6565
6 576 320 256 4811
7 407 247 160 3677
8 324 161 163 2940
9 837 595 242 6608
10 972 517 455 6771
Mean ± Std 617 ± 237 (324–972) 353 ± 177 (120–626) 264 ± 95 (139–455) 4817 ± 1840
(minimum–maximum) (2874–7432)

The last row represents the mean ± standard deviation (minimum–maximum).

Figure 2. Application for skin dose mapping.

Measured vs calculated doses mapping in IC procedures. Since the application uses

DICOM headers to generate a skin dose map, it
Figure 5 shows the difference between the measured
can be implemented for interventional X-ray systems
PSD by the XR-RV3 Gafchromic films and the cal-
from different manufacturers with minimal modifica-
culated PSD by the application for all patients. To
tions in code, as long as all private attributes from
validate the dose calculation tool, deviation was cal-
Table 1 can be found in DICOM header. The appli-
culated between the measured and the calculated PSD
cation was developed with the intention to ensure the
for each patient. Considering all the analyzed pro-
ease of use. User only needs to choose DICOM files
cedures, deviations between the measured and the
generated after the completion of a cardiac procedure
calculated PSD ranged between −37.1 and 43.9%,
and to insert the cumulative dose after which a color-
whereas mean deviations were 8.7 ± 26.3%.
coded skin dose map of 2D patient is generated.
The difference between the measured PSD by XR-
DISCUSSION RV3 Gafchromic films and the calculated PSD using
The originally developed application presented in this our application ranged between −37.1 and 43.9%,
paper is a practical and efficient way for skin dose whereas the mean deviations were 8.7 ± 26.3%.
5
 M. KRAJINOVIĆ ET AL.

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

Figure 3. XR-RV3 Gafchromic calibration curve for determining skin dose.

Figure 4. Geometrical comparison between XR-RV3 Gafchromic films (left) and simulations (right).

6
 SKIN DOSE MAPPING IN INTERVENTIONAL CARDIOLOGY

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

Figure 5. Difference between the measured and the calculated PSD for all patients.

Gafchromic films are used as the gold standard in polymethyl methacrylate (PMMA) phantom using
this study; however, it should be noted that skin Siemens Artis Zee and GE Innova IGS interventional
dose assessment expanded uncertainty using XR-RV3 systems, respectively.
Gafchromic films is 9% (k = 2) for tightly controlled The recognized limitations of the solution pre-
measurement conditions, adequate laboratory cali- sented in this paper are explained in the following
brations and well-defined readout protocol. A more text. Firstly, DICOM header does not contain the
realistic expanded uncertainty is 41% (k = 2), where data regarding the table movement in a 3D space,
a well-defined laboratory calibration is performed, therefore several assumptions are made, including: (1)
while other influencing parameters related to clinical the table does not move in the lateral and longitudinal
application of dosimetry films are less controlled(23) . direction, and (2) the table height is constant, so the
It should also be pointed out that the calculated skin table does not move in vertical direction. Lack of this
dose has non-negligible uncertainty. Only calibration information contributes to the uncertainty of a skin
factor, CF, contributes to expanded uncertainty dose map. DICOM RDSR, however, contains all the
with 17% (k = 2) according to annual quality necessary information regarding the table movement
control tests; therefore, it is safe to assume that skin in 3D space(5) . Secondly, the 3D modeling of the
dose assessment expanded uncertainty using this patient would inevitably represent a more realistic
application is beyond 20% (k = 2). dose assessment, especially for larger angulations and
Since there are no geometric and dosimetric data oblique projections. The size of the patient as well
on each fluoroscopic exposure, it is assumed that flu- as the exact position of the patient on the table can
oroscopic kerma is equally spread on each cine acqui- additionally make the dose assessment more precise.
sition, but in reality, it could be different. DICOM Finally, skin dose map is generated after the proce-
RDSR is a solution to these problems, since it keeps dure; however, it would be more practical to have an
data for each radiation exposure. online skin dose mapping solution.
To the best of our knowledge, only three 2D
applications verified with Gafchromic XR-RV3 films
CONCLUSION
in clinical situations were developed. Greffier et al.(18)
accomplished average difference of 3.4 ± 21.1% This paper presents an originally developed IC
between Gafchromic XR-RV3 films and the appli- skin dose mapping application for many legacy
cation when fluoroscopic kerma was equally spread fluoroscopy units without DICOM RDSR globally
on each cine acquisition. Bordier et al.(9) compared a installed which have not been updated yet, and may
calculation method that produces dose maps using never be. The application uses standardized DICOM
Gafchromic XR-RV3 films on a water phantom headers to generate a skin dose map and represents
and a body phantom and found that dose values a suitable alternative to XR-RV3 Gafchromic film in
agreed within 24.9% accuracy. Habib Geryes et al.(7) monitoring delivered dose during IC procedures. The
accomplished average difference of 10 ± 7% and goal of future research is implementation of DICOM
9 ± 7% between the calculated and the measured RDSR and development of skin dose maps for 3D
PSD values for 34 test conditions performed on human models.

7
 M. KRAJINOVIĆ ET AL.

Downloaded from https://academic.oup.com/rpd/advance-article-abstract/doi/10.1093/rpd/ncaa002/5727935 by Uppsala Universitetsbibliotek user on 20 June 2020

ACKNOWLEDGEMENTS skin during fluoroscopic procedures. Proc. SPIE Int. Soc.
Opt. Eng 8668 (2013).
This work was supported by the Serbian Ministry of 12. Rana, V. K., Rudin, S. and Bednarek, D. R. A track-
Education and Sciences under Grant Agreement No. ing system to calculate patient skin dose in real-time
43009. during neurointerventional procedures using a biplane
X-ray imaging system. Med. Phys. 43(9), 5131–5144
(2016).
REFERENCES 13. Borrego, D., Marshall, T., Tran, D. and Siragusa, W. B.
1. ICRP Publication 120. ICRP Publication 120: radiolog- Physical validation of UF-RIPSA: a rapid in-clinic peak
ical protection in cardiology. Ann. Int. Comm Radiol. skin dose mapping algorithm for fluoroscopically guided
Prot. 42, 1–125 (2013). interventions. J. Appl. Clin. Med. Phys. 19(3), 343–350
2. Ciraj Bjelac, O., Dabin, J., Farah, J., Järvinen, H., (2018).
14. Johnson, P. B., Borrego, D., Balter, S., Johnson, K., Sir-
Malchair, F., Siiskonen, T. and Knežević, Ž. Patient agusa, D. and Bolch, W. E., Eds Skin dose mapping for
maximum skin dose in interventional procedures in radi- fluoroscopically guided interventions. Med. Phys. 38(10),
ology and cardiology: summary of WG 12 activities. 5490–5499 (2011).
EURADOS Report 2019–02. 15. Khodadadegan, Y., Zhang, M., Pavlicek, W., Paden, R.
3. IEC. Medical electrical equipment - Part 2-43: par- G., Chong, B., Schueler, B. A., Fetterly, K. A., Langer,
ticular requirements for the basic safety and essential S. G. and Wu, T. Automatic monitoring of localized skin
performance of X-ray equipment for interventional dose with fluoroscopic and interventional procedures. J.
procedures, Report No. 60601–2–43, 2010. Digit. Imaging 24, 626–639 (2011).
4. Miller, D. L., Balter, S., Dixon, R. G., Nikolic, B., 16. Takata, T., Kotoku, J., Maejima, H., Kumagai, S.,
Bartal, G., Cardella, J. F., Dauer, L. T. and Stecker, M. Arai, N., Kobayashi, T., Shiraishi, K., Yamamoto, M.,
S. Quality improvement guidelines for recording patient Kondo, H. and Furui, S. Fast skin dose estimation
radiation dose in the medical record for fluoroscopically system for interventional radiology. J. Radiat. Res. 59,
guided procedures. J. Vasc. Interv. Radiol. 23, 11–18 1–7 (2017).
(2012). 17. Hellström M. Estimating patient peak skin dose with flu-
5. National Electrical Manufacturers Association. The oroscopic procedures, Umea: Department of Radiation
DICOM standard 2018. Available on https://www.dico Science, Umea University (2018).
mstandard.org/ (accessed 17 October 2019). 18. Greffier, J., Van Ngoc Ty, C., Bonniaud, G., Moliner,
6. Sechopoulos, I., Trianni, A. and Peck, D. The DICOM G., Ledermann, B., Schmutz, L., Cornillet, L., Cayla,
radiation dose structured report: what it is and what it is G., Beregi, J. P. and Pereira, F. Assessment of peak skin
not. J. Am. Coll. Radiol. 12(7), 712–713 (2015). dose in interventional cardiology: a comparison between
7. Habib Geryes, B., Hadid-Beurrier, L., Waryn, M. J., Gafchromic film and dosimetric software em.dose. Phys.
Jeanpierre, A. and Farah, J. Benchmarking the DACS- Med. 38, 16–22 (2017).
integrated radiation dose monitor® skin dose mapping 19. D. Clunie, DICOM encoding of dose information: Stan-
software using XR-RV3 Gafchromic films. Med. Phys. dards, tools and practicalities. 71st Annual Scientific
45(10), 4683–4692 (2018). Congress of JSRT, Yokohama (2015).
8. Bordier, C., Klausz, R. and Desponds, L. Accuracy of 20. International Atomic Energy Agency. Implementation
a dose map method assessed in clinical and anthropomor- of the international code of practice on dosimetry in diag-
phic phantom situations using Gafchromic films. Radiat. nostic radiology. Vienna: International Atomic Energy
Prot. Dosimetry 165(1–4), 244–249 (2015). Agency, (2011).
9. Bordier, C., Klausz, R. and Desponds, L. Patient dose 21. Benmakhlouf, H., Bouchard, H., Fransson, A. and
map indications on interventional X-ray systems and val- Andreo, P. Backscatter factors and mass energy-
idation with Gafchromic® XR-RV3 film. Radiat. Prot. absorption coefficient ratios for diagnostic radiology
Dosimetry 163(3), 306–318 (2015). dosimetry. Phys. Med. Biol. 56(22), 7179–7204 (2011).
10. Bednarek, D. R., Barbaritis, J., Rana, V. K., Nagaraja, 22. Table of mass energy absorption coefficients. Avail-
S. P., Josan, M. S. and Rudin, S. Verification of the able at: https://www.nist.gov/pml/x-ray-mass-attenua
performance accuracy of a real-time skin-dose tracking tion-coefficients (accessed 17 October 2019).
system for interventional fluoroscopic procedures. Proc. 23. Farah, J. et al. Characterization of XR-RV3
SPIE Int. Soc. Opt. Eng. 7961 (2011). GafChromic® films in standard laboratory and in
11. Rana, V. K., Rudin, S. and Bednarek, D. R. Updates in clinical conditions and means to evaluate uncertainties
the real-time dose tracking system (DTS) to improve the and reduce errors. Med. Phys. 42(7), 4211–4226
accuracy in calculating the radiation dose to the patients (2015).