Radiation Protection Dosimetry (2021), Vol. 195, No. 1, pp. 41–49 https://doi.org/10.1093/rpd/ncab111
Advance Access publication 28 July 2021
PATIENT DOSE ASSESSMENT AND OPTIMISATION OF PELVIC
RADIOGRAPHYWITH COMPUTED RADIOGRAPHY SYSTEMS
D. Abbeyquaye1,2,*, S. Inkoom1,3, N.B. Hammond1,4, J.J. Fletcher1,5 and B.O. Botwe6
1Department of Medical Physics, School of Nuclear and Allied Sciences, University of Ghana, Atomic
Energy Campus, P.O. Box AE 1, Atomic Energy-Accra, Ghana
2Department of Biomedical Engineering Technology, Faculty of Health and Allied Sciences, Koforidua
Technical University, P.O. Box KF-981, Koforidua, Ghana
3Radiation Protection and Consultancy Centre, Radiation Protection Institute, Ghana Atomic Energy
Commission, P.O. Box LG 80, Legon-Accra, Ghana
4Department of Nuclear Medicine, National Centre for Radiotherapy and Nuclear Medicine, Korle-Bu
Teaching Hospital, P.O. Box 77, Accra, Ghana
5Department of Applied Physics, Faculty of Applied Sciences, University for Development Studies,
Navrongo Campus, Upper East Region, P.O. Box TL 1350, Tamale, Ghana
6Department of Radiography, School of Biomedical and Allied Health Science, College of Health Sciences,
University of Ghana, P.O. Box KB143, Korle-Bu Campus, Accra, Ghana
*Corresponding author: danny.abbeyquaye@gmail.com
Received 7 December 2019; revised 23 June 2021; editorial decision 7 July 2021; accepted 7 July 2021
Digital radiography systems can reduce radiation dose, this capability was harnessed to explore dose and image quality (IQ)
optimisation strategies. Entrance surface dose (ESD), effective dose (ED) and organ doses were determined by the indirect
method for patients undergoing pelvic anteroposterior X-ray examinations with computed radiography systems. The IQ of
patients’ radiographs was assessed in terms of signal-to-noise ratio (SNR). An anthropomorphic phantom was exposed with
varying tube potential (kVp), tube current-time product (mAs), and focus-to-detector distance (FDD) to determine phantom-
entrance dose for the optimisation studies. SNR of each phantom radiograph was determined. Patients’ mean ESD of
2.38 ± 0.60 mGy, ED of 0.25 ± 0.07 mSv and SNR of 8.5 ± 2.2 were obtained. After optimisation, entrance dose was reduced
by 29.2% with 5 cm increment in FDD, and 5 kVp reduction in tube potential. kVp and/or mAs reduction with an increment in
FDD reduced entrance dose without adversely compromising radiographic-IQ.
INTRODUCTION as a form of investigational level(7). However, the
Pelvic radiography is a commonly requested X- ALARA philosophy suggests that even when the
ray examination that is associated with an effective estimated patient doses fall below recommended
dose (ED) of 0.28 mSv; which is 20 times higher DRLs, a computed radiography (CR) systems’
than the ED associated with chest PA radiographic capability of dose reduction should be harnessed
examinations(1). Physicians request for pelvic radio- to explore procedures to further minimise patient
graphs to assist in diagnosingmedical conditions such dose. This should be achieved without deterioration
as arthritis that affects the hip, inflammation of the in image quality (IQ) of the radiographs, confident
sacrum-ilium joint, pelvic fractures, hip dislocations, for diagnosis.
stiffness of the back or sacroiliac joint and neoplasms. Utilisation of CR in radiography is on a rapid
The radiation protection concept stipulates that ascendancy in Ghana. However, known patient
the net effect of medical radiation exposure must dosimetry in Ghana(8–11) has been reported using
benefit patients and also be optimised(2); that is, be analogue X-ray systems. Also, dosimetry data on
as low as reasonably achievable (ALARA)(3). This patient pelvis anteroposterior (AP) examinations
means that examinations requiring ionising radiation with digital systems(12, 13) is scanty in Ghana, and
should provide the best possible images with the least the optimisation of dose has not been specifically
radiation exposure(4), in order to reduce the chances addressed in the above-cited studies. Amponsah
of radiation-induced carcinogenesis. In radiological et al.(14) and Inkoom et al.(10) reported methods
optimisation, radiation doses in currently performed of dose reduction. Known patient dosimetry with
medical examinations need to be known(5). The next direct digital radiography system was carried out by
step involves comparing the recorded doses with Inkoom et al.(12), which involved pelvic procedures;
diagnostic reference levels (DRLs); a concept that however, no optimisation attempts were indicated.
was introduced by the International Commission A recent study on dose assessment with CR in
on Radiological Protection (ICRP) Report 73(6) Ghana was conducted in 2016 by Ackom et al.(15).
© The Author(s) 2021. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
D. ABBEYQUAYE ET AL
In this study, neither pelvic dose estimation nor dose ≤ 0.100. At 80 kVp and 20 mAs, the deviations of the
optimisation effort was reported(15). Consequently, CRs from collimation accuracy and X-ray/light beam
very little or no optimisation procedures for digital alignment were within the acceptable deviation limits
systems exist in the country. Furthermore, recent of ≤ 1.0 cm and ≤3◦, respectively.
concerns regarding dose creep in digital radiographic The optimisation studies were carried out by using
techniques(16) highlights the need for frequent pelvic an Alderson Rando® female anthropomorphic phan-
dose assessments and optimisation studies with tom, which simulates the radiation scattering proper-
digital systems. The pelvis protects many of the ties of human tissue, and was designed by Radiology
critical organs of the urinary and the reproductive SupportDevices Inc. TheAldersonRando® phantom
systems(17), which requires the realisation of optimal was designed with tissue and skeletal properties con-
gonadal safety of patients. sistent with the reference standards established by the
The aim of this study was to assess the entrance International Commission on Radiation Units and
surface dose (ESD), ED and organ doses to patients Measurements (ICRU) Report No. 44(19, 20) and has
undergoing pelvic X-ray examinations with CR a radiographic appearance (as shown in Figure 1),
systems, and compare the ESD, ED and organ which is comparable to an ideal pelvic radiograph
doses with similar studies and DRLs. Additionally, described by Parker et al.(21)
phantom-dose measurements were carried out aimed
at exploring some dose and IQ optimisation strategies
for the optimum exposure (OE) factors for AP pelvic Patient dosimetry
examinations. ESD, ED and organ doses were estimated for 102
patients (males and females), who were undergoing
pelvis AP X-ray examination. Their ages were from
MATERIALS ANDMETHODS 21 to 80 y, with weights ranging between 59 and 90 kg.
This study was conducted at a public teaching hospi- Patients with critical health conditions were excluded
tal in Ghana. Two CRs in X-ray rooms (Room A and from participating in this study. Non-consenting
B) were used. The CR systems were manufactured patients, numbering up to 32, were also excluded
by Shimadzu Corporation, Japan. Each CR had tube from the study; further contributing to a reduction
model 0.6/1 2P38DE-85. The tubes had a maximum in the sample size. Ethical clearance for this study
tube voltage of 150 kV. At 70 kV, the X-ray tubes had was sought and granted by the hospital’s institutional
an inherent filtration of 1.5 mm-Al. Both CRs were review board, and a written consent was sought from
equipped with automatic exposure control and anti- each participating patient prior to their participation
scatter grid configurations. The cassette reader for the in the study. Before exposure, patient anatomical data
CR in Room A was an Agfa CR cassette reader: CR (such as age, sex, weight and pelvis AP thickness) of
15-X, whereas the CR in Room B was coupled to a each patient were measured. Patients’ weights and
Fujifilm CR reader: FCR Capsula XL II. pelvis thicknesses were measured with a weighing
As part of the quality control (QC) tests, the CRs scale and a physical body calliper, respectively.
used in this study were tested for kVp accuracy, half- Exposure factors that were recorded include kVp,
value layer (HVL) and mAs linearity using Piranha mAs and focus-to-detector distance (FDD). A total
dose meter type 657 to ascertain if each X-ray equip- of 100 cm FDD was used throughout the study
ment was working self consistently. Collimation accu- with respect to standard beam collimation field sizes
racy and beam alignment tests were carried out using relative to the size of patients. The summary of patient
collimation test tool model 161 B and beam align- data and radiographic exposure factors are presented
ment test tool model 162 A. The QC test results in Table 1.
were assessed based on the recommended limits from
the Institute of Physics and Engineering in Medicine
report 91(18). The QC test results showed that the CRs ESD measurement
were in good condition for clinical diagnosis, as each ESD was determined using the indirect method(22–24)
QC test result was within recommended tolerances. of patient dose estimation. A possible limitation with
The maximum tube voltage inaccuracy of both CRs indirect ESD estimates with nomograms is their
at 80 kVp was within the acceptable deviation of ≤ relative less accuracy with respect to direct dose
±6.0%. At 20 mAs and 80 kVp, CR-A and CR-B had estimationmethods. However, indirect ESD estimates
a HVL of 3.62 and 3.36 mm-Al, respectively, these is acceptable for routine dose estimates(22–24). With
were higher than the recommended HVL deviation the indirect method used in this study, X-ray tube
limit of ≥2.10 mm-Al. For the mAs linearity test, the and clinical technique factors as well as patient
maximum difference in mGy/mAs between adjacent anatomical data were recorded and calculated to
stations was 0.010 for CR-A and 0.033 for CR-B; obtain the dose(8) once the appropriate backscatter
which was within the acceptable deviation limit of factor is known. ESD of patients was determined
42
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
PELVIC DOSIMETRY AND OPTIMISATION
Figure 1: Comparison of radiographs after optimisation studies.
using Equation (1)(23): air kerma (INAK) for regularly performed X-ray
[ ( ) ( ) ( ) ] radiographic procedures
(27). AKAP for the ED was
2 estimated by means of normalisation quantities andkV β= p 100 BSFESD mAs , (1) CCs calculated using Monte Carlo methods withα
100 SSD 100 CalDose_X(27). CalDose_X determines organ and
tissue absorbed doses for human phantoms nor-
malised to measurable quantities according to ICRU
where α and β are constants adopted fromHarpen(23) Report 74(28). In line with the concepts outlined in
that depend upon the X-ray generator type, tube that report, CalDose_X presents organ and tissue
voltage in kVp, SSD is the source-to-skin distance in absorbed doses normalised to the AKAP(27). The
cm (equal to FDD—patient pelvic AP thickness), and software estimates AKAP by utilising the energy
BSF represents backscatter factor. According to the dependent fluence-to-air kerma conversion factor(27)
EC, the value of the BSF ranges from 1.2 to 1.4(25). given in ICRP report 74(29). CalDose_X also uses the
However, Martin(24) affirmed using BSF of 1.4 as age, sex and tissue- or organ-specific risk coefficients
the suitable value for ESD/entrance surface air-kerma to estimate the risks of cancer incidence(8, 29–31).
(ESAK) calculations involving pelvic radiography.
Measurement of organ and tissue dose
Estimation of ED
Patient-specific organ doses in the pelvis were
According to the recommendations of ICRP Report calculated with CalDose_X and are presented in
103(26), ED should be a sex-specific weighted dose Table 4. CalDose_X uses voxel phantoms, MAX06
determined from the equivalent doses of reference andFAX06, tomodel an adult patient. The phantoms
tissues and can be calculated using Equation (2)(26): contain the 29 organs and tissues specified by ICRP
Report 103(26), necessary for calculating ED with the
{∑ } tissue weighting factors of ICRP Report 103(26, 27).
= 1 + The software simulation conditions were analogousED WT [HT (female) HT (male)] ,2 to the real patient examination procedure.
(2)
IQ assessment
where HT is the equivalent dose in a tissue or organ, Quantitative IQ assessment of both the patients’
T, and WT is the tissue weighting factor. and the phantom’s radiographs were assessed by
A more desirable approach of measuring of estimating the signal-to-noise ratio (SNR) of the
ED and organ doses with Monte Carlo software radiographs. In this assessment, the properties of the
is by employing direct air-kerma area product digital radiographic regions of interest (ROIs) were
(AKAP) measurement. In this study, ED was selected with consistency on all radiographs, using
determined with Caldose_X 5.0. This is a Monte ImageJ, at locations described by Alzyoud et al.(32)
Carlo-based computer program that takes the tissue and Mraity(33). The ROIs were of the same area
and organ absorbed doses into consideration, as (314.28 mm2) and were drawn on both patient and
well as EDs for posture-specific male and female phantom radiographs to compute pixel mean and
adult phantoms and estimates patient ED using standard deviation values generated with ImageJ
Equation (2)(27). The software also uses conversion computer program. ImageJ is a Java image processing
coefficients (CCs) normalised to ESAK and incident and analysis software that can display, edit and
43
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
D. ABBEYQUAYE ET AL
Table 1. Patient data and radiographic exposure factors at FDD of 100 cm.
Age (years) Tube potential Tube current × Exposure SSD Weight Thickness
(kVp) time (mAs) time (ms) (cm) (kg) (cm)
Range 21–80 68–80 12.5–25.0 25–63 77–77 59–90 21–30
(min–max)
Mean ±σ 54.0±15.0 71.0 ±3.9 19.0 ±3.0 48 ±13.2 74±1.68 71.0 27.0±1.77
±6.62
analyse 8, 16 and 32–bit images in JPEG, DICOM, in DICOM format was assessed for its SNR using
TIFF and other formats(34). SNR is a good predictor ImageJ and with Equation (3).
of IQ(35, 36), which has a positive correlation between
visual grading assessment (VGA)(37) and is widely
used as a measure to judge radiographic IQ(33). It Data analysis
describes the relationship between noise and contrast Statistical analyses were carried out using IBM SPSS
levels in an image for an object with a large scale(38). version 25. As part of the optimisation study, a
The correlation between SNR and the human eye’s one-way analysis of variance (ANOVA) test was
detectability of objects within an image was first conducted to determine the statistical significance
experimentally considered by Albert Rose(39, 40). The between the respective averages of the entrance dose
author found that SNR had a threshold value of
≥ and IQ. This was done by setting a significance5, which is the minimum limit for the human eye threshold of p < 5%. Spearman’s rank-order cor-
to appreciably visualise anatomical structures(39, 41). relation was used to evaluate the correlation strength
Using the radiographic pixel-value characteristics of between the ranks of ESD and ED. A Spearman
eachROI, the SNRof radiographs inDICOMformat correlation coefficient (ρ) of +1 indicates a perfect
was evaluated using Equation (3)(41): positive association of the ESD and ED ranks, and
∑ a ρ of −1 indicates a perfect negative association
(x − x ) between ranks of ESD and ED. A significance level
SNR = i i BG , (3) of p < 5% was used in this test as well.
σBG
where xi is the pixel value in object, xBGis the average RESULTS AND DISCUSSION
pixel in background and σBG is the pixel standard
deviation in the background. Dose assessment
Patient data and exposure factors were recorded from
Optimisation studies the hospital for 56 females and 46 males. The mean
primary radiographic exposure factors used at the
The optimisation study was conducted using an adult hospital for AP pelvic radiography were compared
female Rando anthropomorphic phantom. A total of with exposure factors of other studies fromGhana(8),
24 different exposures were made with varying kVp, Iran(43), UK(44) and Korea(45). The compared param-
mAs and FDD. The kVp ranged from 65 to 75 kVp; eters were the range of patient ages, AP pelvic thick-
the mAs was from 12.5 to 22 mAs, and an FDD ness, SSD, weight and their respective averages with
of 100–105 cm. Combinations of primary exposure standard deviations (σ ) as shown in Table 1. Aver-
factors were selected over a broad range of combi- agely, 71 kVp, 19 mAs at 100 cm FDD was used for
nations that reflect the combinations used clinically AP pelvic examination in the hospital.
at the hospital; and were deemed to be good enough The age and weight range for adults used in this
to produce radiographs of good IQ. The dose to the study was within the adult age range used in the study
phantom was directly measured using Unfors dose by Ofori et al.(8). The mean kVp was similar to the
meter (type 401) and ESAK (mGy) was estimated compared studies(44, 45) with the exception of Iran(43)
using Equation (4)(42) with a BSF of 1.4(24): and Ghana(8). The studies(8, 3) reported a mean tube
voltage of 66 and 77.8 kVp, respectively. Nevertheless,
ESAK = INAK× BSF (4) the kVp range used in this study was within the range
used in the other studies(44, 45). Even though the range
of mAs used in this study was within the range used in
where ESAK is ESAK, INAK is INAK and BSF is other studies, the mean mAs in this study was lower
backscatter factor. IQ of each phantom radiograph than that of the compared studies(8, 43–45).
44
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
PELVIC DOSIMETRY AND OPTIMISATION
Table 2. Estimated patient ESD.
This study Other studies
Mean ±σ range Inkoom Aldrich Iran(43) UK(44) Japan(46) EC(47)
(min–max) et al.(12) et al.(7)
ESD 2.38±0.60 1.53–4.32 6.68 4.75 3.18 3.20 3.00 10.00
(mGy)
Table 3. Estimated patient ED.
This study Other studies
Mean±σ Range (min-max) Aldrich et al.(7) Aliasgharzadeh Vilar-Palop UK(1)
et al.(48) et al.(49)
ED (mSv) 0.25±0.07 0.16–0.45 0.75 0.28 0.37 0.28
The estimated mean ESD and mean ED is pre- doses in this study were compared to the organ doses
sented in Tables 2 and 3, respectively. A Spearman’s presented by Wall et al.(1) for individual radiographs.
correlation between the ranks of the ESD and ED The reported organ doses by Wall et al.(1) for blad-
gave a ρ of 0.915, indicating a very strong positive der, testes and prostate were higher than the doses
correlation between the ranks. A significance level of obtained in this study. However, the organ dose to the
p < 0.001 showed that Spearman’s rank-order corre- ovaries in this studywas higher than the dose reported
lation result is statistically significant. The estimated by Wall et al.(1)
value of the AP pelvis ESD in this study was com- The radiation dose from a single pelvic AP radio-
pared to the ESD of other studies(7, 12) and DRLs(43, graphic procedure was as high as 1.78 mGy for testes
44, 46, 47). Table 2 presents the summary and distribu- and 0.98 mGy for small intestinal wall. Such organs
tion of this comparison. The estimated ESD of 2.38± and tissues are superficial and tend to be in the direct
0.60 mGy was below all compared studies. It was path of the beam. Dose levels reduced to 0.44 mGy
also observed that the ESD in this study was lower for lymphatic nodes and 0.18 mGy for the pancreas.
than that of DRLs from Iran(43), UK(44), Japan(46) Only a small fraction of these organs and tissues lies
and EC(47). Even though the estimated ESD was less in the direct path of the beam, thus receiving lower
than these DRLs, the radiation protection concept dose. In increasing order of dose magnitude, the most
requires that if it is possible andmeasures are available affected reproductive organs are the testes, followed
to further reduce the X-ray dose to patients under- by the ovaries. The most irradiated digestive organs
going diagnostic examination, then those measures are the small intestine, colon and pancreas. The blad-
should be taken. This should be done as long as the der, associated with the urinary system, received a
dose reduction would not compromise the diagnostic mean dose that was >60% of the listed organs. The
integrity of the radiographs. inguinal lymph nodes and the skeletons are relatively
The ED estimated in this study is shown in Table 3. deep lying and receive almost the same dose; higher
The estimated ED was less than all the other ED than that of the pancreas.
values from the compared studies(1, 7, 48, 49). The ED
of 0.25 ±0.07 mSv from this study was similar to the
ED from the study of Aliasgharzadeh et al.(48) and Image quality
that of Wall et al.(1) in the UK; in which, the value The mean SNR values of the IQ assessment of
was obtained using the definitions of ICRP Report patients was 8.50±2.22 and a third quartile value
103(26). of 10.30 was obtained. Patients’ radiographic images
were of good quality. The patient radiograph with the
Radiation dose to organs and tissues minimum SNR had a mean value of 5.80. This meansthat the X-ray images acquired for this study from
The absorbed doses to some pelvic organs and tissues the hospital were all quality enough for confident
are presented in Table 4. Themean dose to the respec- diagnosis, thus dose reduction procedures could be
tive organs and tissues was below the threshold(50, 51) conducted while maintaining acceptable SNR levels.
for the occurrence of deterministic effects. The organ It is recommended that VGA and other physical IQ
45
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
D. ABBEYQUAYE ET AL
Table 4. Mean dose to pelvic organs and tissues.
Organ/tissue Organ/tissue dose (mGy) Organ dose (mGy) Wall et al.(1)
Mean±σ Range (min–max) Mean
Bladder wall 0.80 ±0.32 0.43–1.55 1.30
Colon wall 0.86 ±0.31 0.57–1.55 —
Ovaries 0.69 ±0.29 0.45–1.28 0.52
Small intestines wall 0.98 ±0.36 0.59–1.85 —
Pancreas 0.18 ±0.08 0.08–0.31 —
Skeleton average 0.37 ±0.15 0.20–0.71 —
Lymphatic nodes 0.44 ±0.19 0.20–0.90 —
Testes 1.78 ±0.23 0.93–2.16 2.10
Prostate 0.65 ±0.21 0.31–0.92 0.87
Uterus 0.51 ±0.22 0.33–0.97 —
assessment methods should complement SNR IQ was at 70 kVp, and the last set of exposures was at
estimates to give an overall evaluation of a good 75 kVp. In the first set of exposure, it was observed
clinical image(37). that when the FDDwas increased by 5 cm, the ESAK
The mean SNR of the phantom-optimisation reduced from 0.89 to 0.77 mGy with a corresponding
study was 13.68±1.60 with a third quartile value reduction in SNR from 14.82 to 13.84. When the
of 14.87. The mean SNR of radiographs from the FDD was increased by 5 cm in the second set of
phantom study were 60% higher than that of the exposure, the ESAK reduced from 1.09 to 0.95 mGy
patient study. This could be attributed to exposing and there was a corresponding reduction in SNR
the phantom with similar primary exposure factors from 16.23 to 13.38. Then in the last set of exposure
used for patients, and a limitation arising from the the SNR once again reduced from 14.40 to 13.49
subsequent use of SNR in both instances for IQ when the ESAK reduced from 1.30 to 1.12 mGy at
estimation; that could result in signal fluctuations an FDD increment of 5 cm. In reference to the Rose
resulting from the variations in exposure factors. criterion(39), the reduction in SNR did not render the
Also, there was a notable variation in average images diagnostically irrelevant in all the instances. A
thickness of the patients and phantom (patient one-way ANOVA test done for this analysis showed
= 27 cm and phantom = 20 cm). This results in a statistically significant relation (p = 0.02). It is a
additional attenuation of X-ray beam and increment demonstration that any patient-dose reduction tech-
in scatter radiation(52) in the patients. Another factor niques must be balanced with measures to maintain a
was the minor variation in the procedural techniques good IQ of the resulting radiographs.
arising from image acquisition of different patients by
different radiographers. Observed patient fidgeting
during exposure, introducing motion artefacts(52), OE factors and summary
was also a factor in the relative low SNR of patients From the phantom study, the observed OE factors for
compared to that of the phantom. A limitation in pelvis AP radiography are shown in Table 5. The opti-
the use of the anthropomorphic phantom in IQ mum combination of the kVp, mAs and FDD was
assessment and subsequent comparison with patient selected according to the radiograph that produced
radiographs is indicated. As compared to the vast a good SNR value relative to the other images, and
heterogeneity of human tissues, the phantom’s tissue- had a moderate X-ray dose. The resulting radiograph
substitute material is inherently nearly homogenous for OE is illustrated in Figure 1a. Because the IQ
with no variability in size and shape, thus could resulting from OE is very high, it offered room to
result in under-attenuation of the X-ray beam below explore further dose reduction while maintaining a
90 kV(53). This could result in higher SNR values SNR value above the threshold value set by the Rose
relative to patient SNR values. model(39).
Also, from Table 5, the FDD was increased by
5 cm while maintaining the kVp and mAs settings,
Optimisation this is labelled as ‘ESAK reduction on OE’. The net
photon flux decreased as the inverse of the square
Relationship between dose and IQ of the FDD. Suppose the X-ray photon flux at
The first set of two exposures from the optimisation (FDD)1 =100 cm is 1, and the photon flux at
study was at 65 kVp, the second set of two exposures (FDD)2 =105 cm is 2; then, the 5 cm increment
46
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
PELVIC DOSIMETRY AND OPTIMISATION
Table 5. Exposure factors after optimisation.
Description kVp mAs FDD (cm) ESAK (mGy) SNR
OE 70 12.5 100 1.0865 16.23
ESAK reduction 70 12.5 105 0.9426 13.38
on OE
Further ESAK 65 12.5 105 0.7693 13.04
reduction on OE
in FDD reduced 2 by a factor of ∼ 0.911. It was increase in kVp and a reduction in mAs at a
thus observed that the ESAK reduced by 13.24% and constant FDD resulted in an increment in entrance
the SNR reduced by 17.5% to a SNR value of 13.38, dose. Maintaining kVp and FDD while increasing
which was quality enough for accurate diagnosis. mAs also resulted in dose increment. However, an
The acquired radiograph for this exposure is shown increment in FDD while maintaining kVp and mAs
in Figure 1b. When the kVp was reduced from 70 to settings resulted in lower dose with corresponding
65 kVp (labelled as ‘Further ESAK reduction on OE’ decrease in IQ. The OE factors that were observed
in Table 5), the ESAK reduced by 29.2% as compared after the optimisation study was at 70 kVp, 12.5 mAs
to the dose from the selected OE factors. and 100 cm FDD with ESAK and SNR of 1.09 mGy
Decrease in kVp reduces maximum photon energy and 16.23, respectively. Varying the FDD showed
in the bremsstrahlung spectrum and a reduction in that the ESAK could be reduced by 13.6% with a
photon flux(41, 54). The exposure is approximately 5 cm increment in the FDD.
proportional to the square of kVp(41); thus at 70 Further study aimed at implementing the OE fac-
kVp, 1.16 units of exposure occur for every 1 unit tors in a clinical setting as a patient trial should
of exposure at 65 kVp. A consequential reduction be conducted. Again, further pelvis dosimetry and
in radiographic SNR was indicated due to increase dosimetry for other examinations where radiosen-
in electronic noise(55). The resulting reduction in IQ sitive organs are located in other body parts such
was 19.6% to a SNR of 13.04 (Figure 1c), which was as abdomen, lumber spine, chest and skull regions,
>100% higher than the recommendation by the Rose should be explored to assess options of patient dose
model(37). Hence, the resulting image after the opti- reduction associated with these procedures for the
misation study was of optimum diagnostic quality. In purposes of optimisation. Output exposure, which
Figure 1b and c, a degree of increased unsharpness depends on the tissue penetration effect of different
is observed relative to Figure 1a. This is a geomet- kVp, for the optimisation study should be indicated.
ric unsharpness triggered by making exposures at a
greater FDD (105 cm) for Figure 1b and c, as com-
pared to the FDD of 100 cm at which the exposure
was made for Figure 1a. ACKNOWLEDGEMENTS
Sincere thanks go to all the radiographers at the
KBTH for their assistance in data collection. The sup-
CONCLUSION port of Mr Kofi Ofori (Nuclear Regulatory Author-ity, Ghana) for his help in the use of CalDose_X and
The estimated ESD of 2.38 ± 0.60 mGy for AP some acquisition parameters is acknowledged. The
pelvis radiography was lower than DRLs by at least authors are also grateful to Dr Samuel Nii Adu Tagoe
24%. At 0.25 ± 0.07 mSv the ED when compared to (KBTHRadiotherapyDepartment) for assisting with
other studies was lower by at least 3% in the majority some of the materials used in this work.
of the cases which may be due to the use of good
radiographic techniques at the hospital. Radiographs
of patients were found to be of good quality and REFERENCES
could be used for accurate diagnosis; the mean SNR
of patients’ radiographs was 8.5 ± 2.2, which was 1. Wall, B. F., Haylock, R., Jansen, J. T. M., Hillier, M.
70% higher than the minimum requirement for radio- C., Hart, D. and Shrimpton, P. C. Radiation Risks
graphs to be considered as being relevant for accurate from Medical X-ray Examinations as a Function ofthe Age and Sex of the Patient. Chilton: (Health Pro-
and confident diagnosis. tection Agency Centre for Radiation, Chemical and
The direct influence of exposure factors on Environmental Hazards) (2011).
dose was confirmed. An increase in kVp and/or 2. World Health Organization (WHO). Radiation Protec-
mAs resulted in higher dose with higher IQ. An tion Concepts and Principles. Commun. Radiat. Risks
47
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
D. ABBEYQUAYE ET AL
Paediatr. Imaging. World Heal. Organ., Chapter 2, tems. IPEM Report 91. York Inst. Phys. Eng. Med.
29–48 (2016). (2005).
3. Miller, D. L. and Schauer, D. The ALARA principle in 19. ICRU. Tissue Substitutes in Radiation Dosimetry and
medical imaging. AAPM Newsl. 40, 38–40 (2015). Measurement. International Commission on Radiation
4. Kanal, K. University of Washington Medical Centre Units and Measurements. (1989).
(UWMC). CT Scan dose. UWMC Imaging Services., 20. Radiology Support Devices Inc (RSD). The Alderson
(2016). Available on https://www.uwmedicine.org/si Radiation Therapy Phantom. Available at: http://
tes/default/files/2018-10/CT-Scan-Doses-English.pdf rsdphantoms.com/radiation-therapy/the-alderson-ra
(cited 2021.7.20). diation-therapy-phantom/. (Accessed: 13 February
5. Park, M. Y. and Jung, S. E. Patient dose management: 2020)
focus on practical actions. J. Korean Med. Sci. 31, S45– 21. Parker, S., Nagra, N. S., KulKarni, K., Pegrum, J.,
S54 (2016). Barry, S., Hughes, R. and Ghani, Y. Inadequate pelvic
6. International Commission on Radiological Protec- radiographs: implications of not getting it right the first
tion (ICRP). Radiological Protection and Safety in time. R. Coll. Surg. 99(7), 534–539 (2017).
Medicine ICRP Publication 73. Ann. ICRP 26 (2). 22. Jones, A. K., Heintz, P., Geiser, W., Goldman, L.,
(Oxford: Pergamon Press) (1996). Jerjian, K., Martin, M., Peck, D., Pfeiffer, D., Ranger,
7. Aldrich, J. E., Duran, E., Dunlop, P. and Mayo, J. R. N. and Yorkston, J. Ongoing quality control in digital
Optimization of dose and image quality for computed radiography: report of AAPM Imaging Physics Com-
radiography and digital radiography. J. Digit. Imaging mittee Task Group 151. Med. Phys. 42, 6658–6670
19, 126–131 (2006). (2015).
8. Ofori, K., Gordon, S. W., Akrobortu, E., Ampene, A. 23. Harpen, M. D. A mathematical spread sheet application
A. andDarko, E. O.Estimation of adult patient doses for for production of entrance skin exposure nomograms.
selected X-ray diagnostic examinations. J. Radiat. Res. Am. Assoc. Phys. Med. 23(2), 241–242 (1995).
Appl. Sci. 7, 459–462 (2014). 24. Martin, C. J.Management of patient dose in radiology in
9. Ofori, E. Patient radiation dose assessment in pelvic X- the UK . Radiat. Prot. Dosimetry 147, 355–372 (2011).
ray examination in Ghana. Omi. J. Radiol. 2(8), 2–5 25. European Commission (EC). European Guidelines on
(2013). Quality Criteria for Diagnostic Radiographic Images in
10. Inkoom, S., Schandorf, C. and Fletcher, J. Optimi- Paediatrics. Eur. Comm. (1996).
sation of patient radiation protection in conventional 26. International Commission on Radiological Protection
X-ray imaging procedures using film reject analysis: a (ICRP). ICRP 103: The 2007 Recommendations of the
demonstration of the importance of rare earth screen- International Commission on Radiological Protection.
film systems. Radiatiation Prot. Dosim. 136, 196–202 (Elsevier for ICRP) (2007).
(2009). 27. Kramer, R., Khoury, H. J. and Vieira, J. W. CAL-
11. Schandorf, C. and Tetteh, G. K. Analysis of dose and Dose_X - a software tool for the assessment of organ
dose distribution for patients undergoing selected X- and tissue absorbed doses, effective dose and cancer risks
Ray diagnostic procedures in Ghana. Radiatiation Prot. in diagnostic radiology. Phys. Med. Biol. 53, 6437–6459
Dosim. 76, 249–255 (1998). (2008).
12. Inkoom, S., Togobo, J., Emi-Reynolds, G., Oddoye, A., 28. International Commission On Radiation Units And
Ntiri, T. O. and Gyekye, P. K. Retrospective patient Measurements (ICRU). Patient dosimetry for X-
dose analysis of Ghana’s first direct digital radiography rays used in Medical Imaging. J. ICRU 5(2), iv–vi
system. Health Phys. 103, 133–137 (2012). (2005).
13. Aweligiba, S., Amoako, J. and Schandorf, C. Determi- 29. International Commission on Radiation Units and
nation of Doses and Cancer Risk to Patients Undergoing Measurements (ICRP). Conversion coefficients for use
Digital X-Ray Examinations at the Tamale Teaching in radiological protection against external radiation. Int.
Hospital. University of Ghana (UG) Space (2015). Comm. Radiol. Prot. Ann. ICRP 26(3–4) (1996).
14. Amponsah, E., Boadu, M. and Schandorf, C. Dose 30. Yacoob, H. Y. Assessment of patients X-ray doses at
Reduction in General Radiography While Maintaining three government hospitals in Duhok city lacking require-
Diagnostic Confidence for Selected Examinations in ments of effective quality control. J. Radiat. Res. Appl.
Selected Hospitals in the Ashanti Region, Ghana. Uni- Sci. 10, 183–187 (2017).
versity of Ghana (UG) space (2014). 31. Inyang, S. O., Essien, I. E. and Antia, A. D. Entrance
15. Ackom, D., Sosu, E., Inkoom, S. and Schandorf, C. surface air kerma for chest X-ray examination in some
Estimation of entrance surface dose of adult patients diagnostic radiologic facilities in Akwa Ibom State, Nige-
undergoing computed radiography examinations in two ria. Adv. Phys. Theor. Appl. 48, 4–8 (2015).
hospitals. IJSRSET 3, 746–752 (2017). 32. Alzyoud, K., Hogg, P., Snaith, B., Flintham, K. and
16. Mc Fadden, S., Roding, T., de Vries, G., Benwell, England, A. Radiography impact of body part thickness
M., Bijwaard, H. and Scheurleer, J. Digital imaging on AP pelvis radiographic image quality and effective
and radiographic practise in diagnostic radiography: an dose. Radiography 25, e11–e17 (2019).
overview of current knowledge and practice in Europe. 33. Mraity, H.A.A. B. Optimisation ofRadiation dose and
Radiography 24, 137–141 (2018). image quality for AP pelvis radiographic examination.
17. Tortora,G. J. andDerrickson, B. Principles ofAnatomy (Manchester: University of Salford) (2015).
& Physiology. Hoboken, USA: (John Wiley & Sons, 34. Liu, F., Hernandez-cabronero, M., Sanchez, V., Mar-
Inc.) (2012). cellin, M. W. and Bilgin, A. The Current Role of Image
18. Scally, A. Recommended Standards for the Routine Per- Compression Standards in Medical Imaging. Informa-
formance Testing of Diagnostic X-Ray Imaging Sys- tion 8(4), 1–26 (2017).
48
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021
PELVIC DOSIMETRY AND OPTIMISATION
35. Månsson, L. G. Methods for the evaluation of image Diagnostic reference levels based on latest surveys
quality: a review. Radiat. Prot. Dosimetry 90, 89–99 in Japan – Japan DRLs – 2015.. (2015). Available on
(2000). http://www.radher.jp/J-RIME/report/DRLhoukoku
36. Alves, A. F. F., Alvares, M., Ribeiro, S. M., Duarte, syoEng.pdf (cited 2020.1.28).
S. B., Miranda, J. R. A. and Pina, D. R. Association 47. EuropeanCommission (EC).Diagnostic reference levels
between subjective evaluation and physical parameters in thirty six countries. Radiation protection No. 108.
for radiographic images optimization. Phys. Med. 32, Part 2/2. Publications office of the EU, Luxembourg.
123–132 (2016). (2014).
37. Sandborg, M., Tingberg, A., Ullman, G., Dance, D. 48. Aliasgharzadeh, A., Mihandoost, E., Masoumbeigi,
R. and Carlsson, G. A. Comparison of clinical and M., Salimian, M. and Mohseni, M. Measurement
physical measures of image quality in chest and pelvis of entrance skin dose and calculation of effective
computed radiography at different tube voltages. Med. dose for common diagnostic X-ray examinations
Phys. 33(11), 4169–4175 (2006). in Kashan, Iran. Global J. Health Sci. 7(202–207)
38. Lança, L. and Silva, A.Digital radiography detectors—a (2015).
technical overview: part 2. Sci. Direct , 1–5 (2008). 49. Vilar-Palop, J., Vilar, J., Hernández-Aguado, I.,
39. Rose, A. The Visual Process-Human and Electronic González-Álvarez, I. and Lumbreras, B. Updated
Vision. (New York-London: Plenum Press) (1973). effective doses in radiology. J. Radiol. Prot. 36, 975–990
40. Alsleem, H. and Davidson, R. Quality parameters and (2016).
assessment methods of digital radiography images. Aust. 50. Brenner, D. and Huda, W. Effective dose: a useful con-
Soc. Med. Imaging Radiat. Ther. 59(2), 46–55 (2012). cept in diagnostic radiology? Radiat. Prot. Dosimetry
41. Bushberg, J. T., Seibert, J. A. and Boone, J. M. 128, 503–508 (2008).
The Essential Physics of Medical Imaging-3rd Edi- 51. Wagner, L. K., Eifel, P. J. and Geise, R. A. Potential
tion. (Wolters Kluwer: Lippincott Williams & Wilkins) biological effects following high X-ray dose interven-
(2012). tional procedures. J. Vasc. Interv. Radiol. 5(1), 71–84
42. International Atomic Energy Agency (IAEA). Dosime- (1994).
try in Diagnostic Radiology: an Internacional Code of 52. Al-murshedi, S., Hogg, P. and England, A.Relationship
Practice. Technical Report No. 457. (2014). between body habitus and image quality and radiation
43. Asadinezhad, M. and Bahreyni Toossi, M. T. Doses to dose in chest X-ray examinations: a phantom study. Phys.
patients in some routine diagnostic X-ray examinations Med. 57, 65–71 (2019).
in Iran: proposed the first Iranian diagnostic reference 53. Shrimpton, P. C., Wall, B. F. and Fisher, E. S. The
levels. Radiat. Prot. Dosimetry 132, 409–414 (2008). tissue-equivalence of the Alderson Rando anthropomor-
44. Hart, D., Hillier, M. C. and Shrimpton, P. C. Doses phic phantom for x-rays of diagnostic qualities. Phys.
to Patients from Radiographic and Fluoroscopic X-ray Med. Biol. 26(1), 133–139 (1981).
Imaging Procedures in the UK – 2010 Review. Health 54. Carroll, B. Q. Radiography in the Digital Age. (Illinois,
Protection Agency (UK) - HPA-CRCE-034. (2012). USA: Charles C. Thomas Pub. Ltd) (2018).
45. Kim, Y.-H. et al. Patient dose measurements in diagnos- 55. Cho, H.-M., Kim, H.-J., Choi, Y.-N., Lee, S.-W., Ryu,
tic radiology procedures in Korea. Radiat. Prot. Dosime- H.-J. and Lee, Y.-J. The effects of photon flux on energy
try 123, 540–545 (2007). spectra and imaging characteristics in a photon-counting
46. Yonekura, Y. Japan Network for Research and x-ray detector. Phys. Med. Biol. 58(2013), 4865–4879
Information on Medical Exposure (J-RIME). (2013).
49
Downloaded from https://academic.oup.com/rpd/article/195/1/41/6329435 by University of Ghana. Balme Library user on 14 December 2021