Abstract
Objective
To assess the effect of scan length, partial humeral head inclusion, and body mass index (BMI) on radiation dose of CTs of the medial clavicular epiphysis (MCE) performed with elevated arms (arms-up).
Methods
Ninety consecutive MCE CTs acquired in an arms-up position for forensic age estimation were retrospectively analysed. Scans were grouped by scan length (short: ≤4 cm; long: >4 cm, rounded to the nearest cm) and by arm position (arms-out: humeral heads excluded; arms-in: humeral heads partially included). Dose-relevant parameters (effective tube current, volume CT dose index, CT dose length product, and effective dose) were compared. Adjusted geometric mean ratios (aGMRs) were calculated to quantify the association between radiation doses and scan length, arm inclusion, and BMI.
Results
Population: 88 males, 2 females; median (IQR) weight 62.0 kg (57.5–67.9), height 172 cm (168–178), BMI 21.1 kg/m² (19.5–22.4). Effective dose was lowest in short scans with arms-out (0.38 mSv, 0.32–0.43), highest in long scans with arms-in (0.49 mSv, 0.39–0.62), and equal in short scans with arms-in and long scans with arms-out (0.41 mSv, 0.36–0.46 and 0.34–0.46, respectively). Each additional centimetre of scan length increased dose by 8.3% (aGMR 1.083, 95% CI 1.036–1.132), each centimetre of humeral head inclusion by 6.8% (1.068, 1.046–1.090), and each BMI unit by 4.1% (1.041, 1.024–1.058).
Conclusions
Effective dose was 22% lower in short scans with arms-out than in long scans with arms-in. Careful patient positioning and precise scan length planning substantially reduce radiation exposure in MCE CT for age estimation.
Keywords: Forensic age estimation, Medial clavicular epiphysis, Computed tomography, Radiation protection, Arm positioning
Introduction
The increasing number of potentially unaccompanied minor asylum seekers (PUMAS) arriving in Europe has created a growing demand for forensic age estimation (FAE) to determine eligibility for the legal protections and welfare services granted to minors, including protection from deportation [1–3]. Over the past 25 years, the Arbeitsgemeinschaft für Forensische Altersdiagnostik (AGFAD, Study Group of Forensic Age Diagnostics) has played a central role in standardising FAE procedures: All PUMAS receive radiographs of the teeth and left hand, with a computed tomography (CT) scan of the medial clavicular epiphysis (MCE) added for those with complete epiphyseal fusion of the hand and wrist [4–7].
For non-medical radiological exams like FAE, minimising radiation exposure is essential. MCE CT dose can be reduced using three complementary strategies: low-dose CT protocols [8, 9], positioning the arms outside the beam path [2, 10], and keeping the scan length short [5, 11].
Low-dose CT protocols are already well established in clinical radiology [12] and Gassenmaier et al. demonstrated that their image quality is sufficient for determining the Schmeling main stages [13] and Kellinghaus sub-stages [14] of MCE ossification [8]. Following these results, Gassenmaier recommended a dose limit of 1.0 mSv for MCE CTs [8]. Subsequent work by Tozakidou et al. [2] and Kuhnen et al. [10] showed that this limit is achievable by positioning the arms outside the beam path – in an arms-up position – a standard practice for chest CTs in clinical radiology [15–18]. Kuhnen et al. observed a 33% reduction in radiation dose in arms-up CTs compared to arms-down CTs, even though partial inclusion of the humeral head remained in some arms-up cases due to suboptimal positioning [10]. Because of the small size of the study population, the specific impact of humeral head inclusion on the radiation dose of arms-up CTs could not be quantified [10]. Finally, Mühler et al. and others proposed limiting MCE CT scan length to 4 cm but did not provide quantitative data on the resulting dose reduction [5, 11, 19]. This lack of data represents a notable gap in the literature, given that scan length is a major determinant of CT radiation dose [20, 21].
The primary aim of this retrospective analysis was to determine the effects of scan length, partial humeral head inclusion in the scan beam path, and body-mass index (BMI) on CT radiation dose of MCE CT scans performed with arms-up.
Materials and methods
This retrospective analysis included the CT dose-relevant data of all PUMAS referred to the Department of Diagnostic, Interventional, and Paediatric Radiology at the University Hospital of Bern, Switzerland for forensic age estimation from September 2019 to September 2021.
Scans were performed on a 64 slice CT (Edge, Siemens Healthineers, Erlangen, Germany) with a peak tube voltage of 120 kVp, reference CT tube current of 37 mAs (mAs ref), and automatic dose modulation (care-dose, Siemens Healthineers, Erlangen, Germany). CT scan length was planned on the scout view, using the upper contour of the MCE as landmark. The starting point of the CT scan was 10 mm above the superior contour, the endpoint 10 mm below the inferior contour of the MCE, as measured jointly by the radiographer and reporting radiologist (Fig. 1). All MCE CTs were performed in an arms-up position.
Fig. 1.
Scan length planning. CT scan length planning is performed on the scout view (a). For better delineation of the cortical contours of the clavicle, the scout view is magnified (b). The starting point of the CT scan is 10 mm above the superior contour of the clavicle, the endpoint is 10 mm below the inferior contour of the MCE. In this case, the planned scan length is 42 mm
The field of view was coned to 200 mm for high image resolution. Scans were performed using helical image acquisition with a pitch factor of 0.6. Slice thickness was 1 mm, increment was 0.5 mm, and reconstruction kernels included a bone kernel (I70f\5) and a soft tissue kernel (I31f\3). Image analysis was performed using the hospital’s picture archiving and communication system (PACS) (IDS7, Sectra AB, Linköping, Sweden). Multiplanar reformations (MPR) were used to view true coronal and true axial images of each MCE.
For each case, the following data was retrieved from the PACS: effective CT tube current (mAs eff), volume CT dose index (CTDIvol), and CT dose length product (DLP) (see Table 1). Weight [kg] and height [cm] of the PUMAS was measured prior to CT. The effective CT dose [mSv] was calculated by multiplying the DLP by the organ-specific index for the chest (DLP [mGy cm] × 0.0137 = effective dose [mSv]) [20, 21].
Table 1.
Overview of relevant CT parameters
| CT parameters | Unit1 | Corresponds approximately to |
|---|---|---|
| CT tube voltage | kVp | Energy of x-rays produced by CT x-ray tube |
| Effective CT tube current (mAs eff) | mAs | Amount of x-rays produced by CT x-ray tube |
| Volume CT dose index (CTDIvol) | mGy | Average radiation dose level during scan |
| CT scan length | cm | Length of the body area scanned |
| CT dose length product (DLP) | mGy*cm | Total radiation output for the scan |
| Effective CT dose | mSv | Estimated radiation exposure of the patient |
1kVp = kilo-Voltage peak; mAs = milli-Amperes per second; mGy = milli-Grey; cm = centimetres; mSv = milli-Sievert
All PUMAS were positioned on the CT table with arms raised above the head. Due to individual radiographer practice, hand positioning varied and included hands parallel to the head, hands crossed at the wrists above the head, or hands placed under the head. Scout view images were retrospectively reviewed to assess whether one or both humeral heads were included within the scan beam. When humeral head inclusion was present, the length of inclusion within the scan beam was measured on the scout view (cm) separately for each arm and as the combined sum of both arms (Fig. 2). The effective CT scan length was measured on coronal CT images (cm).
Fig. 2.
Measurements of humeral head inclusion. Comparison of three scout view images, all with arms raised above the head. Optimal arm position with both humeral heads outside of the scan beam path (a). Suboptimal arm position with partial inclusion of the left humeral head in the scan beam (b) and nearly complete inclusion of both humeral heads in the scan beam path (c)correctly positioned
All CTs were retrospectively grouped according to scan length and arm position: Scans with partial humeral head inclusion in the beam path were classified as arms-in, those with complete exclusion of the arms as arms-out. Scan length was measured on the scout view and rounded to the nearest centimetre. Scans rounded to ≤ 4 cm were classified as short scans (corresponding to a maximum true length of 4.4 cm), while those rounded to > 4 cm were classified as long scans (corresponding to a minimum true length of 4.5 cm). This approach ensured stable group sizes, accounted for minor variations in actual scan length due to table speed inaccuracies or automatic overranging to reconstruct the first and last images [22], and reflected clinical practice, where small deviations of ± 0.5 cm are considered acceptable and fall within the expected range of measurement inaccuracy.
Statistical analysis
The statistical analysis was performed with STATA 18 (StataCorp, College Station, Texas, USA). Continuous variables are presented as median with interquartile ranges (IQRs). P-values were calculated using for group comparisons using Wilcoxon rank test for two group comparisons and Kruskal-Wallis test for more than two group comparison. Categorical variables were reported as counts and percentages (%), with p-values obtained from either Chi-squared test or a Fisher’s exact test, as appropriate. A p-value of < 0.05 was used as threshold for statistical significance. A multivariable linear regression analysis was performed with ln-transformed effective CT dose as the dependent variable and scan length, arm position, and BMI as independent factors. The exponentiated regression coefficients were used to derive adjusted geometric mean ratios (aGMRs) with a confidence interval of 95% to quantify the relationships between effective CT doses and scan length, partial arm inclusion, and BMI, respectively [23].
Results
Population characteristics
90 consecutive CTs of the MCE were evaluated, all scanned with arms elevated above the head. There were 88 (98%) male and 2 (2%) female PUMAS. Median weight, height and body mass index (BMI) were 62.0 kg (57.5–67.9), 172.0 cm (168.0–177.5), 21.1 kg/m2 (19.5–22.4).
Scan characteristics
Median scan length of all 90 cases was 4.9 cm (4.2–5.4). Scan length was ≤ 4 cm in 26 cases (29%, median 4.1 cm, 3.9–4.2) and > 4 cm in 64 cases (71%, 5.1 cm, 4.8–5.7) (Table 2).
Table 2.
Effect of scan length on radiation dose
| Patient characteristics | All scans (n = 90) | Scan length ≤ 4 cm (n = 26) | Scan length > 4 cm (n = 64) | P-value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Body mass index (kg/m2) | 90 | 21.1 | (19.5; 22.4) | 26 | 21.2 | (19.9; 22.4) | 64 | 21.0 | (19.4; 22.5) | 0.621 |
| Weight (kg) | 90 | 62.0 | (57.5; 67.9) | 26 | 62.2 | (58.0; 68.0) | 64 | 62.0 | (57.3; 67.8) | 0.793 |
| Height (cm) | 90 | 172.0 | (168.0; 177.5) | 26 | 170.5 | (166.0; 177.5) | 64 | 173.0 | (169.0; 177.2) | 0.207 |
| Scan characteristics | ||||||||||
| CT scan length (cm) | 90 | 4.9 | (4.2; 5.4) | 26 | 4.1 | (3.9; 4.2) | 64 | 5.1 | (4.8; 5.7) | < 0.001 |
| Total humeral head inclusion (R&L) (cm) | 90 | 0.4 | (0.0; 2.3) | 26 | 0.3 | 90.0; 1.1) | 64 | 0.1 | (0.0; 2.8) | 0.169 |
| Right humeral head inclusion (cm) | 90 | 0.0 | (0.0; 1.2) | 26 | 0.0 | (0.0; 0.4) | 64 | 0.4 | (0.0; 1.7) | 0.028 |
| Left humerus head inclusion (cm) | 90 | 0.5 | (0.0; 1.0) | 26 | 0.5 | (0.0; 0.4) | 64 | 0.1 | (0.0; 1.2) | 0.368 |
| Humeral head position, n (%) | 90 | 26 | 64 | |||||||
| Arms in | 55 | (61.1) | 16 | (61.5) | 39 | (60.9) | ||||
| Arms out | 35 | (38.9) | 10 | (38.5) | 25 | (39.1) | 0.958 | |||
| Scan length/humeral head inclusion | 90 | 26 | 64 | |||||||
| SL: ≤4 cm, Arms out | 10 | (11.1) | 10 | (38.5) | 0 | (0.0) | ||||
| SL: ≤4 cm, Arms in | 16 | (17.8) | 16 | (61.5) | 0 | (0.0) | ||||
| SL: >4 cm, Arms out | 25 | (27.8) | 0 | (0.0) | 25 | (39.1) | < 0.001 | |||
| SL: >4 cm, Arms in | 39 | (43.3) | 0 | (0.0) | 39 | (60.9) | ||||
| Radiation doses | ||||||||||
| Effective CT tube current (mAs) | 90 | 65.5 | (58.0; 81.0) | 26 | 69.0 | (59.0; 83.0) | 64 | 64.5 | (58.0; 78.0) | 0.325 |
| CT dose index (mGy) | 90 | 4.5 | (3.9; 5.5) | 26 | 4.7 | (4.0; 5.6) | 64 | 4.3 | (3.9; 5.3) | 0.306 |
| Dose length product (mGy*cm) | 90 | 30.0 | (27.0; 38.0) | 26 | 28.5 | (24.0; 32.0) | 64 | 32.0 | (28.0; 41.0) | 0.007 |
| Effective dose (mSv) | 90 | 0.4 | (0.4; 0.5) | 26 | 0.4 | (0.3; 0.4) | 64 | 0.4 | (0.4; 0.6) | 0.007 |
Continuous variables are described with median (IQR), p-values obtained by Wilcoxon rank sum test
Categorical variables are shown with number (%) in each category, p-values obtained by Chi-squared test
Fisher’s exact test was additionally calculated if possible (otherwise signed by -), when the expected frequency of one cell was less than 5
The arms were completely excluded from the scan in 35 cases (39%), while one or both humeral heads were partially included in 55 cases (61%). Median left and right humeral head inclusion was 0.4 cm (0.0–2.3) for all scans and 1.8 cm (0.8–3.6) for the 55 cases with partial humeral head inclusion (Table 3). Among the 26 CTs with a short scan length of ≤ 4 cm, 10 were scanned with arms-out and 16 with partial humeral head inclusion. Among the 64 scans with a long scan length of > 4 cm, 25 were scanned with arms-out and 39 with partial humeral head inclusion (Tables 2 and 3).
Table 3.
Effect of partial humeral head inclusion on radiation dose
| Patient characteristics | All scans (n = 90) | Scans arms out (n = 35) | Scans arms in (= 55) | P-value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Body mass index (kg/m2) | 90 | 21.1 | (19.5; 22.4) | 55 | 21.0 | (19.6; 22.1) | 35 | 21.6 | (19.0; 23.0) | 0.634 |
| Weight (kg) | 90 | 62.0 | (57.5; 67.9) | 55 | 61.2 | (56.9; 67.5) | 35 | 62.7 | (58.2; 69.0) | 0.317 |
| Height (cm) | 90 | 172.0 | (168.0; 177.5) | 55 | 171.0 | (167.0; 177.0) | 35 | 174.0 | (170.0; 178.0) | 0.081 |
| Scan characteristics | ||||||||||
| CT scan length (cm) | 90 | 4.9 | (4.2; 5.4) | 35 | 4.9 | (4.2; 5.2) | 55 | 4.8 | (4.4; 5.5) | 0.904 |
| Scan length group, n (%) | 90 | 35 | 55 | |||||||
| ≤ 4 cm | 26 | (28.9) | 10 | (28.6) | 16 | (29.1) | ||||
| > 4 cm | 64 | (71.1) | 25 | (71.4) | 39 | (70.9) | 0.958 | |||
| Total humeral head inclusion (R&L) (cm) | 90 | 0.4 | (0.0; 2.3) | 35 | 0.0 | (0.0; 0.0) | 55 | 1.8 | (0.8; 3.60 | < 0.001 |
| Right humeral head inclusion (cm) | 90 | 0.0 | (0.0; 1.2) | 35 | 0.0 | (0.0; 0.0) | 55 | 0.9 | (0.2; 1.9) | < 0.001 |
| Left humerus head inclusion (cm) | 90 | 0.1 | (0.0; 1.0) | 35 | 0.0 | (0.0; 0.0) | 55 | 0.9 | (0.3; 1.70 | < 0.001 |
| Humeral head position, n (%) | 90 | 35 | 55 | |||||||
| Arms in | 55 | (61.1) | 0 | (0.0) | 55 | (100.0) | ||||
| Arms out | 35 | (38.9) | 35 | (100.0) | 0 | (0.0) | < 0.001 | |||
| Humeral head inclusion/scan length, n (%) | 90 | 35 | 55 | |||||||
| Arms out, SL: ≤4 cm | 10 | (11.1) | 10 | (28.6) | 0 | (0.0) | ||||
| Arms in, SL: ≤4 cm | 16 | (17.8) | 0 | (0.0) | 16 | (29.1) | ||||
| Arms out, SL: >4 cm | 25 | (27.8) | 25 | (71.4) | 0 | (0.0) | < 0.001 | |||
| Arms in, SL: >4 cm | 39 | (43.3) | 0 | (0.0) | 39 | (70.9) | ||||
| Radiation doses | ||||||||||
| Effective CT tube current (mAs) | 90 | 65.5 | (58.0; 81.0) | 35 | 58.0 | (50.0; 74.0) | 55 | 68.0 | (61.0; 84.0) | < 0.001 |
| CT dose index (mGy) | 90 | 4.5 | (3.9; 5.5) | 35 | 3.9 | (3.4; 5.0) | 55 | 4.7 | (4.2; 5.7) | 0.001 |
| Dose length product (mGy*cm) | 90 | 30.0 | (27.0; 38.0) | 35 | 29.0 | (24.0; 32.0) | 55 | 33.0 | (28.0; 42.0) | 0.003 |
| Effective dose (mSv) | 90 | 0.4 | (0.4; 0.5) | 35 | 0.4 | (0.3; 0.4) | 55 | 0.5 | (0.4; 0.6) | 0.003 |
Continuous variables are described with median (IQR), p-values obtained by Wilcoxon rank sum test
Categorical variables are shown with number (%) in each category, p-values obtained by Chi-squared test or by Fisher’s exact test if appropriate
Effect of scan length on radiation dose
Median DLP and effective CT dose were 30 mGy*cm (27.0–38.0) and 0.4 mSv (0.4–0.5). DLP and effective CT dose were significantly lower in ≤ 4 cm scans than in > 4 cm scans. DLP ≤ 4 cm: 28.5 mGy*cm (24.0–32.0) vs. >4 cm: 32.0 mGy*cm (28.0–41.0); p = 0.007. Effective CT dose ≤ 4 cm: 0.4 mSv (0.3–0.4) vs. >4 cm: 0.4 mSv (0.4–0.6); p = 0.007. BMI, effective CT tube current, and CTDIvol were not significantly different between short and long scans (Table 2).
Effect of partial arm inclusion on radiation dose
Median effective CT tube current, CTDIvol, total DLP, and effective CT dose were significantly lower in scans with complete exclusion of the arms from the scan beam path than in scans with partial humeral head inclusion. Tube current arms-out: 58.0 mAs (50.0–74.0) vs. arms-in: 68.0 mAs (61.0–84.0); p < 0.001. CTDIvol arms-out: 3.9 mGy (3.4–5.0.4.0) vs. arms-in: 4.7 mGy (4.2–5.7); p = 0.001. DLP arms-out: 29.0 mGycm (24.0–32.0) vs. arms-in: 33.0 mGycm (28.0–42.0); p = 0.003. Effective CT dose arms-out: 0.4 mSv (0.3–0.4) vs. 0.5 mSv (0.4–0.6); p = 0.003. There was no significant difference between scans with arms-out and arms-in regarding BMI, and scan length (Table 3).
Combined effect of scan length and partial arm inclusion on radiation dose
Median effective CT dose was highest in > 4 cm scans with arms-in (0.49 mSv (0.39–0.62)) and lowest in ≤ 4 cm scans with arms-out (0.38 mSv (0.32–0.43), p = 0.003) corresponding to a 22% dose reduction. Median effective CT dose was identical in ≤ 4 cm arms-in and > 4 cm arms-out scans (0.41 mSv (0.36–0.46) and 0.41 mSv (0.35–0.46), p = 0.963). >4 cm arms-in scans had significantly higher doses than both > 4 cm arms-out scans (p = 0.003) and ≤ 4 cm arms-in scans (p = 0.004) (Fig. 3, Table 4).
Fig. 3.
Effective dose values of all four scan categories. Comparison of the ranges and median effective dose values for all four categories. Median CT dose was lowest in ≤ 4 cm scans with arms-out (0.38 mSv, category 1, blue boxplot) and highest for > 4 cm scans with arms-in (0.49 mSv). The effective CT dose for ≤ 4 cm scans with arms-in and > 4 cm scans with arms-out was identical (0.41 mSv, categories 2 and 3, red and green boxplots). Median effective CT dose was significantly higher in > 4 cm scans with arms-in than in ≤ 4 cm scans with arms-out (p = 0.003), ≤ 4 cm scans with arms-in (p = 0.004), and > 4 cm scans with arms-out (p = 0.003). There was no significant difference in radiation dose between the categories 1 (blue), 2 (red), and 3 (green)
Table 4.
Arm position, scan length and effective radiation dose
| Arm position | Scan length | N | Median radiation dose |
|---|---|---|---|
| arms-out | ≤4 cm | 10 | 0.38 mSv (0.32–0.43) |
| >4 cm | 24 | 0.41 mSv (0.35–0.46) | |
| arms-in | ≤4 cm | 16 | 0.41 mSv (0.36–0.46) |
| >4 cm | 39 | 0.49 mSv arm (0.39–0.62) |
N: number of cases. Units: cm = centimetres; mSv = millisievert
Effect of scan length, partial arm inclusion, and BMI on radiation dose
The aGMR was 1.083 (95% CI: 1.036–1.132) for scan length, 1.068 (95% CI: 1.046–1.090) for partial arm inclusion and 1.041 (95% CI: 1.024–1.058) for BMI, which reflects an increase in effective dose of about 8.3% for each centimetre increase in CT scan length, of about 6.8% for each centimetre of arm inclusion, and of about 4.1% for each unit increase in BMI (kg/m²).
Discussion
This study found that the effective radiation dose of MCE CTs with short scans (≤ 4 cm) and arms-out was 22% lower than in long scans (> 4 cm) with partial humeral head inclusion. Effective dose increased by approximately 8.3% per additional centimetre of scan length, 6.8% per centimetre of humeral head inclusion, and 4.1% per unit increase in BMI.
These findings are relevant because they demonstrate how careful patient positioning and scan-length planning of MCE CTs can substantially reduce the effective CT radiation dose. Under optimal conditions with a scan length of ≤ 4 cm and complete exclusion of the arms from the scan beam path, the effective dose of MCE CTs was reduced to 0.38 mSv. Compared with previous studies, this represents a further dose reduction of 52% from 0.79 mSv reported by Tozakidou et al. and 37% from 0.6 mSv reported by Kuhnen et al. [2, 10].
While earlier work by had already examined the effect of arms-up versus arms-down positioning on radiation dose, this study additionally investigated the impact of partial humeral head inclusion in MCE CTs performed with arms up [2, 10]. Partial inclusion of one or both humeral heads was observed in 61% of all cases; complete exclusion of the arms form the beam path was achieved in only 39% of scans. This high proportion of partial inclusion was primarily attributable to suboptimal positioning, with arms placed parallel to the head – allowing inadvertent movement between positioning and scanning – instead of being crossed over the vertex with hands holding the elbows (Fig. 4). Optimising arm positioning requires good communication and collaboration between radiologists and radiographers, along with clear, universally understandable instructions for examinees, irrespective of language barriers.
Fig. 4.
Examples of arm positioning during CT examination (a) Arms positioned parallel to the head: this allows for potential arm movement between positioning and scanning and may lead to inclusion of the humeral heads in the scan range. (b) Arms crossed above the vertex with hands holding the contralateral elbows: this position minimizes arm movement and prevents the inclusion of the humeral heads in the scan beam
Although the relative dose increase per additional centimetre was higher for scan length (8.3% per cm) than for humeral head inclusion (6.8% per cm), humeral head inclusion had a greater overall impact on effective dose in this study. The reason is that the median length of humeral head inclusion among the 55 arms-in cases was 1.8 cm (equivalent to 0.9 cm for each arm, or up to 1.8 cm if only one arm was included), whereas the median excessive scan length was only 1.0 cm. Consequently, humeral head inclusion increased the effective dose by approximately 13.6% (6.8% × 1.8 cm), compared with 8.3% (8.3% × 1.0 cm) for excessive scan length. These results underline the importance of correct arm positioning, ensuring that the humeral heads are excluded from the scan beam path. This finding is consistent with established clinical practice of positioning the arms outside the scan region to reduce radiation dose [15–18, 24, 25].
Nevertheless, keeping the scan length as short as possible remains crucial for dose optimisation. Mühler’s proposition to limit the scan length to ≤ 4 cm, made 20 years ago [5], remains valid but appears difficult to achieve in practice: In this study, the median scan length was 4.9 cm – nearly 1 cm longer than the targeted 4 cm. Only 26 of 90 MCE CTs (29%) achieved a scan length ≤ 4 cm. Previous studies likewise reported mean scan lengths exceeding 4 cm, with values of 6.2 ± 2.1 cm and 4.9 ± 0.8 cm in arms-up scans, and 8.5 ± 3.4 cm and 5.2 ± 0.9 cm in arms-down scans [2, 10]. These results indicate that CT scan length planning still requires refinement. A more consistent way to achieve the 4 cm target may be to define the scan length from 10 mm above the superior contour of the MCE to 30 mm below the superior contour (on the scout view), rather than from 10 mm above the superior contour to 10 mm below the inferior contour of the MCE.
CT dose is also influenced by BMI, with higher radiation doses observed in patients with higher BMI [26]. This study found that radiation dose increases by approximately 4.1% per unit increase in BMI. For example, if the CT dose of an MCE CT for a normal-weight individual (BMI 22.2) is 0.35 mSv, the dose for an underweight individual (BMI 17.4) would decrease by 17.6% to 0.29 mSv, while for an overweight individual (BMI 25.1), it would increase by 12.5% to 0.39 mSv. Given the substantial impact of BMI on radiation dose, careful scan length planning and optimal arm positioning are especially important for PUMAS with a high BMI.
Limitations
This study has several methodological limitations. The primary limitation is its single-centre, single-scanner design, which restricts the generalisability of the findings. CT scanners from different generations or manufacturers may yield varying radiation dose values, even when using identical scan parameters. The authors hope to investigate whether these dose reduction strategies have a comparable effect across different CT scanners.
It may be criticised that the upper limit of short-scans was set at 4.4 cm rather than the 4.0 cm originally proposed by Mühler [5]. This adjustment was made to accommodate minor inaccuracies in scan planning and acquisition. As stated in the Methods section, small discrepancies between planned and actual scan length can occur due to technical factors such as table speed variation or overranging, which refers to the automatic extension of the scan beyond the planned endpoint to collect sufficient data for reconstructing the first and last image slices [22]. In addition, measuring scan length on scout views to the nearest millimetre is challenging, and deviations a few millimetres from the planned scan length are common. However, while deviations of are negligible for whole-body trauma CTs (approximately 160 cm scan length) or chest CTs (approximately 50 cm scan length), they may have a significant impact on the dose of short MCE CTs.
A third limitation is the relatively small number of cases in the arms-out group. Although scan length was determined jointly by a radiologist and a radiographer based on scout views, variations in positioning technique or landmark selection could introduce variability in dose optimisation.
Finally, while scan length optimisation and arm positioning are crucial for dose reduction, this study did not assess the influence of tube voltage and tube current on effective dose. Data from post-mortem CT studies suggest that with ultra-low tube voltage and tube current settings, the effective dose for MCE CTs could potentially be reduced to approximately 0.2 mSv [9]. However, such strategies must be carefully evaluated to ensure that diagnostic accuracy is not compromised.
Conclusions
This study examined the effect of CT scan length and arm position in MCE CTs for FAE performed in an arms-up position and found that the effective radiation dose was 22% lower in CTs with a scan length of ≤ 4 cm and arms completely outside the CT beam path than in CTs with a scan length of > 4 cm and partial humeral head inclusion. Overall, radiation dose increased by 8.3% per cm scan length, 6.8% per cm humeral head inclusion, and 4.1% per BMI unit.
These findings underscore the importance of careful arm positioning and scan length planning for radiation protection – especially in radiologic imaging conducted for non-medical reasons in young individuals.
Author contributions
Conceptualisation: Thomas D Ruder, Marc Alexander Schmid, Rainer J Egli. Methodology: Thomas D Ruder, Martin Müller. Data curation: Marc Alexander Schmid, Martin Müller, Wolf-Dieter Zech, Thomas D Ruder. Data analysis: Martin Müller, Thomas D Ruder. Writing – original draft: Thomas D Ruder, Marc Alexander Schmid. Writing – review and editing: Paolo Lombardo, Rebecca Straessle, Andreas Schmeling, Wolf-Dieter Zech, Marc Alexander Schmid, Martin Müller, Rainer J Egli, Thomas D Ruder. Visualisation: Martin Müller, Thomas D Ruder. Supervision: Martin Müller.
Funding
This study received no external funding.
Declarations
Ethical approval
The study was classified as quality evaluation study by the Ethics Committee of the Canton of Bern (KEK Req-2023-00390) and the need for informed consent was waived.
Clinical trial number
not applicable.
Conflict of interest
The authors declare that they have no conflict of interest.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Asylum applicants by (2025) type, citizenship, age and sex - annual aggregated data. https://ec.europa.eu/eurostat/databrowser/view/migr_asyappctza/default/table?lang=en. Accessed 22 Oct
- 2.Tozakidou M, Meister RL, Well L et al (2021) CT of the medial clavicular epiphysis for forensic age estimation: hands up? Int J Legal Med 135:1581–1587. 10.1007/s00414-021-02516-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schmeling A, Dettmeyer R, Rudolf E et al (2016) Forensic Age Estimation: Methods, Certainty, and the Law. Deutsches Ärzteblatt international. 10.3238/arztebl.2016.0044 [DOI] [PMC free article] [PubMed]
- 4.Wittschieber D, Ottow C, Vieth V et al (2015) Projection radiography of the clavicle: still recommendable for forensic age diagnostics in living individuals? Int J Legal Med 129:187–193. 10.1007/s00414-014-1067-0 [DOI] [PubMed] [Google Scholar]
- 5.Mühler M, Schulz R, Schmidt S et al (2006) The influence of slice thickness on assessment of clavicle ossification in forensic age diagnostics. Int J Legal Med 120:15–17. 10.1007/s00414-005-0010-9 [DOI] [PubMed] [Google Scholar]
- 6.Ruder TD, Kuhnen SC, Zech W-D et al (2023) Standards of practice in forensic age estimation with CT of the medial clavicular epiphysis—a systematic review. Int J Legal Med 137:1757–1766. 10.1007/s00414-023-03061-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Schmeling A, Grundmann C, Fuhrmann A et al (2008) Criteria for age estimation in living individuals. Int J Legal Med 122:457–460. 10.1007/s00414-008-0254-2 [DOI] [PubMed] [Google Scholar]
- 8.Gassenmaier S, Schaefer JF, Nikolaou K et al (2020) Forensic age estimation in living adolescents with CT imaging of the clavicula—impact of low-dose scanning on readers’ confidence. Eur Radiol 30:6645–6652. 10.1007/s00330-020-07079-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mehnert S, Berger N, Flach PM et al (2023) How low can we go? CT dose reduction in the assessment of the medial clavicular epiphysis in forensic age estimation: a prospective postmortem CT study. Acta Radiol 64:2126–2131. 10.1177/02841851231164996 [DOI] [PubMed] [Google Scholar]
- 10.Kuhnen SC, Müller M, Schmeling A et al (2025) CT of the medial clavicular epiphysis for forensic age estimation – raised arms position recommended. Int J Legal Med. 10.1007/s00414-025-03521-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lossois M, Cyteval C, Baccino E, Peyron P-A (2022) Forensic age assessments of alleged unaccompanied minors at the medicolegal Institute of Montpellier: a 4-year retrospective study. Int J Legal Med 136:853–859. 10.1007/s00414-022-02813-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lam S, Bai C, Baldwin DR et al (2024) Current and future perspectives on computed tomography screening for lung cancer: a roadmap from 2023 to 2027 from the International Association for the Study of Lung Cancer. J Thorac Oncol 19:36–51. 10.1016/j.jtho.2023.07.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Schmeling A, Schulz R, Reisinger W et al (2004) Studies on the time frame for ossification of the medial clavicular epiphyseal cartilage in conventional radiography. Int J Legal Med 118:5–8. 10.1007/s00414-003-0404-5 [DOI] [PubMed] [Google Scholar]
- 14.Kellinghaus M, Schulz R, Vieth V et al (2010) Forensic age estimation in living subjects based on the ossification status of the medial clavicular epiphysis as revealed by thin-slice multidetector computed tomography. Int J Legal Med 124:149–154. 10.1007/s00414-009-0398-8 [DOI] [PubMed] [Google Scholar]
- 15.Brink M, De Lange F, Oostveen LJ et al (2008) Arm raising at exposure-controlled multidetector trauma CT of thoracoabdominal region: higher image quality, lower radiation dose. Radiology 249:661–670. 10.1148/radiol.2492080169 [DOI] [PubMed] [Google Scholar]
- 16.Euler A, Szücs-Farkas Z, Schindera S (2014) Möglichkeiten der Strahlenreduktion bei der CT des Körperstamms. Radiologie up2date 14:163–176. 10.1055/s-0034-1365687 [Google Scholar]
- 17.Bayer J, Pache G, Strohm PC et al (2011) Influence of arm positioning on radiation dose for whole body computed tomography in trauma patients. J Trauma Injury Infect Crit Care 70:900–905. 10.1097/ta.0b013e3181edc80e [DOI] [PubMed] [Google Scholar]
- 18.Karlo C, Gnannt R, Frauenfelder T et al (2011) Whole-body CT in polytrauma patients: effect of arm positioning on thoracic and abdominal image quality. Emerg Radiol 18:285–293. 10.1007/s10140-011-0948-5 [DOI] [PubMed] [Google Scholar]
- 19.Schulz R, Mühler M, Reisinger W et al (2008) Radiographic staging of ossification of the medial clavicular epiphysis. Int J Legal Med 122:55–58. 10.1007/s00414-007-0210-6 [DOI] [PubMed] [Google Scholar]
- 20.Zhao A, Fopma S, Agrawal R (2022) Demystifying the CT radiation dose sheet. Radiographics 42:1239–1250. 10.1148/rg.210107 [DOI] [PubMed] [Google Scholar]
- 21.Huda W (2010) Review of Radiologic Physics. Lippincott Williams & Wilkins [Google Scholar]
- 22.Schilham A, Van Der Molen AJ, Prokop M, De Jong HW (2010) Overranging at multisection CT: an underestimated source of excess radiation exposure. Radiographics 30:1057–1067. 10.1148/rg.304095167 [DOI] [PubMed] [Google Scholar]
- 23.Friedrich JO, Adhikari NKJ, Beyene J (2012) Ratio of geometric means to analyze continuous outcomes in meta-analysis: comparison to mean differences and ratio of arithmetic means using empiric data and simulation. Stat Med 31:1857–1886. 10.1002/sim.4501 [DOI] [PubMed] [Google Scholar]
- 24.Marsh RM, Silosky MS (2017) The effects of patient positioning when interpreting CT dose metrics: a Phantom study. Med Phys 44:1514–1524. 10.1002/mp.12137 [DOI] [PubMed] [Google Scholar]
- 25.Tozakidou M, Yang S-R, Kovacs BK et al (2019) Dose-optimized computed tomography of the cervical spine in patients with shoulder pull-down: is image quality comparable with a standard dose protocol in an emergency setting? Eur J Radiol 120:108655. 10.1016/j.ejrad.2019.108655 [DOI] [PubMed] [Google Scholar]
- 26.Sperry BW, Vamenta MS, Gunta SP et al (2024) Influence of body mass index on radiation exposure across imaging modalities in the evaluation of chest pain. J Am Heart Assoc. 10.1161/jaha.123.033566 [DOI] [PMC free article] [PubMed] [Google Scholar]