4.2 The results of the patient study
4.2.5 Entrance skin dose
4.2.7.12 Total image score
Figure 18: The distribution of evaluation scores for visualisation of pelvic/hip soft tissues in groups divided by tissue removal
4.2.7.12 Total image score
The median of total image score in the group without tissue removal was at 33.5 (IQR=6.3) and in the group in which patients removed tissue was at 37.0, IQR was 3.4 (p=0.004). The minimum score in the first group was 21.3 and in the second group was 30.7. The maximum score in the first and second group was 40 and 39.7, respectively. The rest of the statistical analysis is shown in Table 25.
Table 25: Statistical analysis of total image score in groups with and without fat tissue removal
Group n Mean ± SD. Median Min. Max. p-value Without tissue removal 30 33.0 ± 4.9 33.5 21.3 40.0 0.004
With tissue removal 30 36.4 ± 2.3 37.0 30.7 39.7
Based on the results of the Mann-Whitney U test, there were statistically significant differences in the total image score between groups (p=0.004). The comparison of the distribution of total image scores in each group is shown in Figure 19.
Figure 19: The distribution of total image scores in groups divided by tissue removal
44
5 DISCUSSION
Pelvic radiography in the erect position provides more diagnostically useful information on functional anatomy and is more suitable for the evaluation of hip pathologies such as DDH and FAI when compared to radiography in a supine position (Flintham et al., 2021; Alzyoud et al., 2018; Jackson et al., 2016; Pullen et al., 2014; Fuchs-Winkelmann et al., 2008).
Radiography of the pelvis in erect position is associated with poorer image quality and higher radiation dose in larger patients compared to those acquired supine (Alzyoud, 2019). This is because soft tissue descends due to gravity, resulting in increased AP thickness in the pelvic region, resulting in poor image quality and increased dose (Flintham et al., 2017).
Given that erect pelvic radiography is so diagnostically useful, we investigated the radiation dose and image quality of pelvic images acquired in erect position while patients manually displaced soft tissue from the pelvic region during the examination. We compared the quality of the images obtained and the radiation doses received between patients who moved the fat tissue from the region of interest (pelvis) and those who did not. In particular, we investigated how moving the fatty tissue during erect pelvic imaging affected the DAP, ESD, effective dose, and image quality.
In the phantom study, which compared fat displacement bands made of four different materials, we found that the thin triangular cotton bandage was best for removing adipose tissue, because it did not produce artefacts and adequately mobilised the fat.
Most patients (56.7%) had class I obesity according to the BMI classification, 36.7% were overweight and 6.7% had class II obesity. We found no statistically significant differences (p=0.790) between patients who had to remove fatty tissue during the study and those who did not. This is important because it means that all outcomes were comparable between these two groups. If one of the groups had a higher BMI than the other, our results would not be comparable in terms of dose calculations, because as Zalokar and colleagues (2020) explained, patients with a higher BMI receive a dose several times higher than patients with normal body weight.
The mean ± SD waist circumference in the group of patients without fat tissue removal was 112.9 ± 7.4 cm and in the group of patients who had soft tissue removed from the region of interest, 111.7 ± 5.3 cm. There were no statistically significant differences in waist circumference between these two groups (p=0.459), indicating that the results between the
groups were comparable. We also found no statistical differences in hip circumference measurements between the groups (p=0.077). The mean ± SD hip circumference for patients who did not have adipose tissue removed was 108.7 ± 7.2 cm and for the other group 105.6 ± 6.1 cm.
When comparing waist and hip circumference measurements before and after fat tissue removal in a group of patients with fat tissue removal, we found statistically significant differences (p<0.001) in waist circumference. When patients manually moved fat tissue from the region of interest using a thin cotton bandage, the thickness around the waist decreased by 4.7% as the shifted tissue moved upward, over the region of interest. Shifting soft tissue during the examination did not affect hip circumference, because fat tissue was not that inferior in most patients, so we did not find any differences in hip circumference measurements before and after fat tissue displacement (p=0.211).
We found no statistically significant differences in primary field size between group of patients who moved the fat tissue and those that did not (p=0.743). Field size was measured at the image receptor surface and mean ± SD 1600.6 ± 144.6 cm2 in patients without tissue removal and 1604.3 ± 113.4 cm2 in patients with fat tissue removal. The measurements did not differ because we had to include the entire region of interest (pelvis) in the field size, so moving the fat tissue did not affect the collimation.
The mean ± SD DAP in patients who did not displace fat tissue was 194.4 ± 106.6 μGy m2 and in patients who removed fat tissue it was 119.8 ± 43.2 μGy m2. We found statistically significant differences between the groups (p=0.001), as DAP was 38.5% lower in a group of patients with fat tissue displacement. The reason for this result is that with the displacement of soft tissue, the body thickness decreases, as previously mentioned by 4.7%
in the waist area, resulting in a lower DAP value for patients in this group. Alzyoud (2019) came to a similar conclusion when comparing AP pelvic projections in the supine and erect AP pelvic projections. After changing from supine to erect position in pelvic radiography, AP thickness increased by 13% in normal patients, 24% in overweight patients, and 19% in obese patients, which also leads to an increase in DAP values by 42%, 55%, and 105% in normal, overweight, and obese patients, respectively.
We found statistical differences in ESD between the two groups of patients (p<0.001). The ESD decreased by 44% in the soft tissue removal group of patients as thickness decreased
46
around the waist. The mean ± SD ESD in a group of patients without fat tissue removal was 5.12 ± 2.68 mGy and in the other group was 2.87 ± 1.05 mGy. Comparing the average values of ESD with the values measured by Aliasgharzadeh and colleagues (2015), the value of ESD (2.87 ± 1.05 mGy) measured in this study in a group of patients with tissue removal is similar to the values in the above-mentioned study in which they investigated the values of ESD during supine pelvic radiography in the AP projection. The ESD measurements in four different hospitals are 1.21, 2.89, 2.70 and 2.80 mGy. The mean ± SD entrance skin dose we measured in patients who did not remove fat tissue from the region of interest (5.12 ± 2.68 mGy) is higher comparing to the research of Aliasgharzadeh et al. (2015), as they measured only in the supine position, where soft tissue shifts laterally and the thickness of the patient decreases. When the pelvis is imaged in the erect position, the tissue decreases due to gravity and the thickness increases and so does ESD. Therefore, the values of ESD in a group with tissue removal are more comparable to the study of Aliasgharzadeh et al. (2015), than to those taken without tissue removal.
Effective dose was reduced by 38.7% in a group of patients who manually displaced fat tissue during imaging. The results were statistically significant (p<0.001). A study by Alzyoud (2019) reported an increase in effective dose when changing from the supine to erect position during AP pelvic radiography. As body thickness increased in the erect position, the effective dose also increased by 38%, 65%, and 120% in the normal, overweight, and obese BMI groups, respectively. In the phantom study by Alzyoud and colleagues (2019), the effective dose increased by 856% at 15 cm fat thickness and 80 kV.
In our study, the mean ± SD effective dose was 237.8 ± 113.2 μSv in the group without fat removal and 145.8 ± 46.1 μSv in the group without tissue removal. Comparing the effective doses in our study with the effective doses of pelvic radiography in Slovenia and other European countries, we found that the effective doses in our study were lower than in the publication European Union, 2014, where the average effective dose for pelvic radiography in Slovenia was 0.52 mSv and in other European countries 0.71 mSv, with a range between 0.21 and 2 mSv (European Union, 2014).
The results of the calculation of the degree of agreement between the radiologists in the evaluation of each criterion in the obtained images show that all three radiologists generally had a low degree of agreement, since the coefficient for each criterion did not exceed the value of 0.40 (good degree of agreement). Radiologists 2 and 3 had the highest level of
agreement, agreeing most on the assessment of visualisation of the hip joints, as the Kappa coefficient showed fair level of agreement (κ=0.321). There was also fair level of agreement in assessing visualisation of the trochanters (κ=0.301), sacroiliac joints (κ=0.254), iliac crests (κ=0.233), and acetabula (κ=0.274). They showed no agreement in the evaluation of the visualisation of the pelvic/hip soft tissues (κ=-0.042).
Radiologists 1 and 2 fairly agreed when evaluating the visualisation of the trochanters (κ=0.274) and had no agreement in evaluation of visualisation of the medulla and cortex of the pelvis (κ=-0.001). Radiologists 1 and 3 had fair level of agreement in assessing visualisation of the trochanters and iliac crests but disagreed when assessing visualisation of the femoral necks (κ=-0.032), sacrum and its foramina (κ=-0.006), and soft tissues of the pelvis/hip (κ=-0.050).
We found statistically significant differences in the assessment of visualisation of the hip joints between the groups with and without fat tissue removal (p=0.001). Hip joints were more visible in the group of patients who removed tissue during the examination, as the median score was 4.0 for this group and 3.7 for the group without tissue removal. There were also significant differences (p=0.021) between groups in trochanter visibility scores. The trochanters were more visible on images with fat tissue removal, with a median score of 3.7 and a median score of 3.2 in a group without tissue removal. Visualisation of the acetabula was rated with a median score of 4.0 in the group where patients removed fatty tissue and 3.7 in the other group. There were statistically significant differences between the groups as the p-value was below 0.001, from which we conclude that the acetabula is more visible in patients with tissue removal. We also found statistical differences in the visualisation of the femoral neck (p=0.021), as they were more visible on the images obtained with soft tissue removal (the median score was 4.0, while in the other group the median score was 3.7).
Visualisation of the medulla and cortex of the pelvis was rated with a median value of 3.7 in the group with soft tissue removal and 3.3 in the group without tissue removal. We found statistically significant differences between these two groups, as medulla and cortex were better visualised in patients who moved the fat tissue (p=0.009). There were statistical differences in the visualisation of the sacrum and its foramina (p=0.008), as they were better visible on the images with tissue removal (the median value was 3.3). The median score for the other group was lower (2.7). We found better visualisation of the pelvic/hip soft tissues in patients with fat tissue removal, as the median value in this group was 3.5, while the
48
median score in the group without tissue removal was 3.0. The results were statistically significant (p=0.039). The reason for the better visualisation of the aforementioned anatomical structures in patients who have moved the fat tissue during imaging is due to the body thickness around the waist, because when body fat increases, the attenuation of the body fat increases, the attenuation of the X-ray beam leads to increased noise and lower contrast resolution, which is why the results were lower in the group of patients who had not removed the fat tissue (Modica et al., 2011).
For sacroiliac joints image score evaluation, we found no statistical differences between patients in both groups, and the p-value was just above the limit of statistical significance (p=0.055). The median score for patients who had moved the fat tissue was 3.7 and for the other group 3.2. The reason that visualisation of the sacroiliac joints did not differ between the groups is that the patients who removed the fat tissue from region of interest moved the tissue in the upper region, i.e., at the level of the sacroiliac joints, so that the thickness in this area increased, resulting in lower contrast resolution and higher noise. This is also the reason why we did not find statistically significant differences in the visualisation of the iliac crests (p=0.060), although the median score was 4.0 in the patients who had removed the fat tissue and 3.7 in those who had not. It is also important to note that the p-value was in the borderline range of statistical significance. We found no statistical differences in the visualisation of the pubic/ischial rami (p=0.166), with a median score at 4.0 in the group of patients who removed fat tissue and 3.7 in the other group.
The median of total image score in the patients with fat tissue removal was higher (37.0) than in the group of patients who did not have soft tissue removed during imaging (33.5).
There were statistically significant differences between the groups in the total image score (p=0.004). This means that image quality was generally better when patients had fat tissue removed from the region of interest, as body thickness decreased, resulting in better contrast resolution and lower noise compared to images obtained by patients without fat tissue removal. Alzyoud (2019) came to the same conclusion when comparing supine and erect images in patients with different BMI. She found that image quality decreased by 6% in normal weight patients, 10% in overweight patients, and 15% in obese patients when they switched from the supine to erect position.
The main difference in the radiographic technique recommended by European Commission, 2004 was that we used lateral chambers instead of the central chamber in our phantom and
patient study. The lateral chambers were used because Kim et al. (2015) reported in their study that the lowest dose is generally measured when lateral chambers are chosen. The recommended criteria (European Commission, 2004) were developed in a film-screen era and were based on the opinion of an expert committee. We have found no other recent recommendations that address the limitation from this publication.
Our study has several limitations. Our sample size was small, only 60 patients with 30 patients in each group. Another limitation is that we used a CR system for the phantom study.
It would have been more appropriate if we had tested the bands with the DR system, since the patient study was conducted on DR system. We also used the PCXMC 2.0 software to calculate the effective dose, but this program does not offer the possibility to select the position in which the radiograph was taken (supine or erect) but only allows the selection of the projection (AP or PA), so the tissue movement due to gravity was not fully taken into account.
50
6 CONCLUSION
In this study we found that when taking a pelvic radiograph in the erect position, while moving fat tissue from the region of interest, the patient dose area product, the entrance skin dose and the effective dose decrease. This also affects the image quality, as it increases with the removal of the tissue and most of the anatomical structures of the pelvis are better visualised.
The waist circumference of the patients decreased by 4.7% when the adipose tissue was removed from the region of interest. This also affected the radiation dose quantities and image quality. DAP was 38.5% lower in a group of patients with tissue removal, also ESD decreased by 44% and effective dose by 38.7%. Image quality was also generally better in the group of patients who had displaced fat tissue removed from the pelvic area, as the total image score was higher than in the other group. The hip joints, trochanters, acetabula, femoral neck, medulla and cortex of the pelvis, sacrum and its foramina, and pelvic/hip soft tissues were better visualised on images after removal of adipose tissue. When evaluating the visualisation of sacroiliac joints, iliac crests, and pubic/ischial rami, we found no statistically significant differences between patients who had fatty tissue removed and those who did not.
The reason for the lower radiation dose and better image quality when removing fatty tissue from the region of interest is that body thickness, i.e., waist circumference, decreases. As the body thickness decreases, a lower dose is required for projection through the pelvis, which means that the patient receives a lower radiation dose. In terms of image quality, the attenuation of the X-ray beam with increasing body thickness leads to increased noise and lower contrast resolution.
Based on the above, the authors recommend that pelvic radiographs are taken in the erect position in overweight and obese patients, removing fatty tissue with a band that does not cause artefacts on the image and removes fatty tissue from the region of interest, so that image quality improves, and the received effective dose is as low as reasonably achievable.
7 REFERENCES
Ahmad N (2003). Proper positioning for the pelvis and proximal femur. Available at:
https://www.auntminnie.com/index.aspx?sec=ser&sub=def&pag=dis&ItemID=58880
<20. 3. 2021>.
Aliasgharzadeh A, Mihandoost E, Masoumbeigi M, Salimian M, Mohseni M (2015).
Measurement of Entrance Skin Dose and Calculation of Effective Dose for Common Diagnostic X-Ray Examinations in Kashan, Iran. Glob J Health Sci 7 (5): 202–7.
doi: 10.5539/gjhs.v7n5p202.
Alzyoud K, Hogg P, Snaith B, Flintham K, England A (2018). Optimum Positioning for Anteroposterior Pelvis Radiography: A Literature Review. J Med Imaging Radiat Sci 49 (3): 316–24. doi: 10.1016/j.jmir.2018.04.025.
Alzyoud K, Hogg P, Snaith B, Flintham K, England A (2019). Impact of body part
thickness on AP pelvis radiographic image quality and effective dose. Radiography 25 (1):
e11–7. doi: 10.1016/j.radi.2018.09.001.
Alzyoud K (2019). Establishing an Evidence-Base for Erect Pelvis Radiography:
Positioning, Radiation Dose and Image Quality. PhD thesis. Manchester: University of Salford.
Campbell SE (2005). Radiography of the hip: Lines, Signs, and Patterns of Disease. Semin Roentgenol 40 (3): 290–319. doi: 10.1053/j.ro.2005.01.016.
Chan VO, McDermott S, Buckley O et al. (2012). The relationship of body mass index and abdominal fat on the radiation dose received during routine computed tomographic
imaging of the abdomen and pelvis. Can Assoc Radiol J 63 (4): 260–6.
doi: 10.1016/j.carj.2011.02.006.
Chiron P, Demoulin L, Wytrykowski K, Cavaignac E, Reina N, Murgier J (2017).
Radiation dose and magnification in pelvic X-ray: EOSTM imaging system versus plain radiographs. Orthop Traumatol Surg Res 103 (8): 1155–9. doi: 10.1016/j.otsr.2017.07.018.
52
Drake RL, Vogl WA, Mitchell AWL (2010). Gray’s anatomy for students. 2nd ed.
Philadelphia: Churchill Livingstone.
Durani Y (2021). X-Ray Exam: Pelvis. Available at: https://kidshealth.org/en/parents/xray-pelvis.html <24. 1. 2021>.
Efthymiou FO, Metaxas VI, Dimitroukas CP, Panayiotakis GS (2020). Low BMI patient dose in digital radiography. Radiat Prot Dosimetry 189 (1): 1–12.
doi: 10.1093/rpd/ncaa007.
European Commission (2004). DIMOND III – Image Quality and Dose Management For Digital Radiography. Available at: https://www.sorf.fi/doc/diamond_III.pdf
<20. 11. 2019>.
European Union (2014). Radiation protection N°180. Medical Radiation Exposure of the European Population: Part 1/2. Luxembourg: Publications Office of the European Union.
Available at: https://ec.europa.eu/energy/sites/ener/files/documents/RP180.pdf
<12. 3. 2021>.
Fisher DR, Fahey FH (2017). Appropriate use of effective dose in radiation protection and risk assessment. Health Phys 113 (2): 102–9. doi: 10.1097/HP.0000000000000674.
Flintham K, Alzyoud K, England A, Hogg P, Snaith B (2021). Comparing the supine and erect pelvis radiographic examinations: an evaluation of anatomy, image quality and radiation dose. Br J Radiol 94 (1123): 1–17. doi: 10.1259/bjr.20210047.
Flintham K, Alzyoud K, England A, Hogg P, Snaith B (2021). Comparing the supine and erect pelvis radiographic examinations: an evaluation of anatomy, image quality and radiation dose. Br J Radiol 94 (1123): 1–17. doi: 10.1259/bjr.20210047.