Forensic Science International 234 (2014) 7–12

Contents lists available at ScienceDirect

Forensic Science International journal homepage: www.elsevier.com/locate/forsciint

Forensic age estimation from the clavicle using 1.0 T MRI—Preliminary results Sara Tangmose a,*, Karl Erik Jensen b, Chiara Villa a, Niels Lynnerup a a b

Section of Forensic Pathology, Department of Forensic Medicine, University of Copenhagen, Frederik V Vej 11, DK-2100 Copenhagen Ø, Denmark Department of Radiology, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej 3, DK-2100 Copenhagen Ø, Denmark

A R T I C L E I N F O

A B S T R A C T

Article history: Received 28 February 2013 Received in revised form 19 August 2013 Accepted 17 October 2013 Available online 31 October 2013

Objectives: As forensic age estimations in the living are performed without medical indication, there is a need for the development of non-ionizing methods. This study investigates the use of 1.0 T MRI to visualize the ossification status of the medial end of the clavicle. Material and methods: T2 weighted 3D images were collected from a 1.0 T MR system. We prospectively scanned 102 subjects, 47 autopsy cases and 55 living volunteers (12–33 years). Images were scored in blind trials by three observers using a 4-stage system. Observers differed by level of training and radiological expertise. Results: Motion artefacts reduced image resolution in living subjects. However, mean age at stage 4 was significantly different from mean age at stage 2 and 3. The minimum age at stage 4 was 19.8 years. Interobserver agreement was moderate for autopsy cases and slight for living subjects. By contrasting subjects of ‘‘incomplete fusion’’ ( stage 3 on both sides) against ‘‘fused’’ (at least one stage 4), agreement rate rose to moderate (K = 0.414) for living subjects. Despite the low agreement rate, no subject younger than 18 years was assessed as having ‘‘fused’’ clavicles. Conclusion: At lower image resolution, a 2-stage system increases agreements rates among observers. To further increase accuracy, clavicle staging needs to be performed by trained observers. If available, a 1.0 T MR system may be used for age estimation in the living. However, further studies are needed to ensure that the ability to discriminate adults from minors, i.e. 18 years, is maintained. ß 2013 Elsevier Ireland Ltd. All rights reserved.

Keywords: Magnetic resonance imaging Clavicle ossification Age estimation Forensic radiology

1. Introduction The ossification status of the medial end of the clavicle is used for forensic age estimation in post-pubertal adolescents and young adults, as it is the last bone to ossify [1]. Clavicle ossification varies greatly and hence, high accuracy age estimates cannot be obtained. Still, the covered age ranges make this bone particularly important concerning the legal age limits 18 and 21 years. At 18 years the hand ossification, third molar mineralisation and sexual maturation should be completed [2]. According to the recommendations of the study group on forensic age estimation (AGFAD), an assessment of the clavicle ossification should be performed when skeletal maturation is completed in the bones of the left handwrist [3]. Well-known methods for assessing the developmental stage of the clavicles include conventional radiography (CR) and computer tomography (CT) [4,5]. At an effective radiation dose of 0.6–0.8 mSv, CT invovles a higher level of radiation compared to CR

* Corresponding author. Tel.: +45 35326170; fax: +45 35326150. E-mail addresses: [email protected], [email protected] (S. Tangmose), [email protected] (C. Villa), [email protected] (N. Lynnerup). 0379-0738/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.forsciint.2013.10.027

(0.2 mSv per chest radiograph) [6]. However, in approximately 9– 13% of plain chest radiographs, a stage cannot be assessed due to superposition of bony structures [7,8]. This percentage may drop to 2.5% using an oblique position [2]. Unfortunately, every radiation exposure increases the theoretical lifetime risk of mortality from radiation induced cancer [9]. Whitout a medical indication, use of radiation is prohibited in some countries [2]. Forenisic age estimations are performed whitout medical indication in young individuals. Thus, the development of non-ionizing methods is important. MRI has shown promising results and a method for high field MRI (3.0 T) has been published [2,10]. However, expensive cost and a limited access may delimit its use [11]. MRI scanners of lower magnetic field strength (1.0 or 1.5 T) are widely available and may serve as a cheaper alternative in Forensic departments as well as a supplement to CT prior to medico-legal autopsies. MR image resolution is in part affected by the signal-to-noise ratio (SNR), motion artefacts, noise, and voxel size. At lower magnetic field, SNR is almost halved as compared to 3.0 T [12]. Reduced SNR may increase slice thickness and/or create longer examination time [13]. Slice thickness affects the accuracy of clavicle staging while examination time affects the ability to use the method in the living [10,14]. While lower magnetic field

8

S. Tangmose et al. / Forensic Science International 234 (2014) 7–12

strength may be well suited for autopsy casework, their use may be limited in the living. As they are situated in a subcutaneous position and close to the heart and lungs, clavicles are prone to noise generated from blood flow, breathing motion and the surface noise arising at the transit from a barrel shaped torso to a slender neck. Thus, at lower field strength, image resolution may be too poor to distinguish the developmental stages of the medial end of the clavicle. The aim of this study was to investigate whether images of the clavicles using a 1.0 T MR system can be applied as a method for forensic age estimation. 2. Material and methods This prospective study included living volunteers as well as autopsy cases of both sex during a three-year period (2010–2012). A total of 102 subjects (74 M, 28 F) between 12.9 and 33.5 years were included comprising 47 corpses and 55 living volunteers (Fig. 1). The corpses submitted to a medico-legal autopsy (35 M, 12 F) were MR scanned in the supine position as part of the routine investigation. Cases of severe decomposition or trauma to the chest area were excluded. Ethnic origin was not taken into account. Healthy volunteers (39 M, 16 F) without previous trauma to the clavicles were scanned using the same MR system and protocol. Of these, 2 females were excluded due to image artefacts. Of the remaining 53, 32 were scanned in the prone position to further reduce breathing artefacts. The local ethics committee approved this part of the study and written informed consent was obtained from all participants and their parents in case of minors. The ethnic origin was mainly European Caucasian. Four individuals originated from the Middle East, Asia or Africa. All were raised in a Nordic country. MR images were collected on a 1.0 T Siemens Harmony1 scanner, Siemens AG, Germany with standard software using a 3D T2 weighted gradient Echo sequence. The evaluation of the ossifying epiphysis required a MR pulse sequence that provided optimal tissue contrast between fluid, cartilage and bone. All the

available pulse sequences were tested on the selected surface coil (receiver only). The test images were evaluated for tissue contrast and SNR in order to provide the best image quality. The technical parameters were: Flip angle 25, slice thickness 1.5 mm, FOV 180 mm, TR: 26.44 ms, TE: 7.29 ms, Scan time 6.04 min, Matrix 256. This gave an in plane resolution and voxel size of 1.5  0.7  0.7 mm. Images were analysed using multi planar reconstruction (MPR), which allowed a simultaneously view of the three dimensions (coronal, transversal and sagittal). Images were assessed in blind trials by three observers. One observer (ST) was a medical doctor with training in age estimation from the clavicle using MRI and CT. The second observer (KEJ) was a radiologist specialized in musculoskeletal MRI, but with no prior training in estimating age from the clavicle. The third observer (CV) was a forensic anthropologist with training in several skeletal age estimation methods other than the clavicle using CT imaging. We used a 4stage system as suggested by Schmidt et al. [10]. Apart from a short introduction, KEJ and CV were self-taught and used the images and stage classifications as described by the paper. All subjects were assessed by ST and CV and 95 were assessed by KEJ. In order to determine the intra-observer agreement, one of us (ST) reassessed all images 3 months after the first reading. Observers analysed images by use of the Osirix Dicom Viewer on high quality monitors. First, clavicles were analysed separately yielding a maximum of two stages from each subject. As analysed simultaneously, they were however, not independent. The relationship between age and stage as assessed among observers were analysed by box plots and one-way ANOVA. Left-right differences were analysed using a Wilcoxon signed rank test and partial correlations, controlling for age. Sex differences were analysed using Mann Whitney U-test. Spearman rank correlations tested stages with age. Secondly, the intra- and inter-observer agreement in autopsy cases as well as living subjects was analysed using Kappa statistics and the related definitions by Landis and Koch [15]. Statistical analysis was performed using SPSS 19.0 software package for statistical analysis

Fig. 1. The distribution of included subjects by sex and group.

S. Tangmose et al. / Forensic Science International 234 (2014) 7–12

9

Fig. 2. Image resolution is reduced in living subjects due to motion artefacts. MRI images representing stage 1, 2, 3 and 4 are shown for living ((A)–(D)) and autopsy cases ((E)– (H)).

(IBM). Unless stated otherwise, a p-value below 0.05 was considered statistically significant. 3. Results Image resolution was reduced in living subjects as compared to autopsy cases due to increased motion artefacts (Fig. 2). One observer (KEJ) assessed 4 clavicles as ‘‘unable to score’’ due to technical artefacts. Of these, 2 originated from the same subject. As seen from the box plots, clavicle scores 2, 3, and 4 tended to increase with age by the three observers (Fig. 3). Using one-way ANOVA, there was a significant effect of the 4 stages on age by each observer [F(3,196) = 49.6, p < 0.001 (ST), F(3,182) = 11.3, p < 0.001 (KEJ), and F(3,196) = 28.0, p < 0.001 (CV)]. Post hoc comparisons using Bonferroni adjustment indicated that mean age at stage 4 was significantly different from mean age at stage 1, 2 and 3 when assessed by ST (p < 0.001) and CV (p < 0.005) and from mean age at stage 2 and 3 when assessed by KEJ (p < 0.001), respectively. By one observer (ST), mean age at stage 2 was different from mean age at stage 3 (p < 0.001).

Fig. 3. Box-plot showing the distribution of chronological age in years by a 4-stage system as (1, 2).

S. Tangmose et al. / Forensic Science International 234 (2014) 7–12

10

Table 1 The mean (standard deviation), minimum and maximum age in years by stage 2, 3, and 4 as assessed by observers as well as reference data from a study using 3.0 T MRI [18]. Stage 1 was excluded due to a low number of cases. M: males, F: females. Observer

ST KEJ CV Hilleweg et al. [19]

Stage 2

Stage 3

Stage 4

N

Mean + SD (yrs.)

Min–max (yrs.)

N

Mean + SD (yrs.)

Min–max (yrs.)

N

Mean + SD (yrs.)

Min–max (yrs.)

M:31 F:19 M:57 F:17 M:28 F:9 M:0 F:1

20.8 18.1 22.5 20.2 20.9 19.1 – –

17.0–24.7 12.9–21.7 17.0–33.5 16.5–27.4 17.0–24.7 12.9–24.6 – –

M:77 F:18 M:29 F:15 M:52 F:27 M:57 F:60

22.7 23.1 22.1 21.4 22.1 21.3 20.6 19.8

19.3–31.2 19.5–29.0 19.3–26.9 12.9–30.7 17.7-–28.8 16.9–30.7 16.2–26.2 16.0–26.0

M:34 F:13 M:37 F:9 M:46 F:16 M:34 F: 44

26.3 26.4 24.3 26.8 25.5 25.3 24.9 24.2

21.1–33.5 21.9–30.7 19.8–31.2 20.6–30.7 19.8–33.5 20.6–29.3 22.1–26.9 18.1–26.9

(2.3) (2.4) (3.3) (2.9) (1.9) (4.5)

(2.2) (2.9) (1.7) (4.8) (2.2) (3.6)

(3.1) (3.1) (2.9) (3.8) (3.2) (3.3)

Table 2 The intra- and inter-observer agreement as assessed in each subject by the three observers. Kappa values contrasts the collapsed composite stages of stage 1, 2 and 3 against at least one stage 4. Observer

Autopsy cases

ST vs. ST ST vs. KEJ ST vs. CV KEJ vs. CV *

Living volunteers

N

Kappa

% Equal

N

47 45 47 45

0.752* 0.567* 0.594* 0.524*

89.4 82.2 83.0 84.4

53 49 53 49

Kappa 0.669* 0.009 0.414* 0.004

% Equal 86.8 63.3 75.5 57.1

Indicate a significant level of agreement.

The percentages of clavicles assessed as stage 4 were 23.5% (47/ 200), 24.7% (46/186) and 31.0% (62/200) by ST, KEJ and CV, respectively. At stage 4, the minimum age was 19.8 years. Table 1 describes the mean, minimum and maximum age by stage (2–4), sex and observers in relation to reference data from a previous MRI study (3.0 T). Overall mean age between the male (23.1 years, SD 3.1) and the female (22.2 years, SD 4.3) sample was not significantly different (p = 0.104, independent samples t-test). Using Mann Whitney Utest, stage assessments by the three observers did not differ between male and female cases (0.121 < p< 0.542). Hence, in the following, sexes were analysed together. Looking bilaterally, a few subjects were assessed with divergent stages. For example, 4 subjects were assessed as a stage 1 on one side and a 2 on the other (1 + 2), 10 subjects as stage 2 + 3 and 5 subjects as stage 3 + 4 by observer ST. One observer (KEJ) assessed the left clavicle at a higher stage than the right (p = 0.03, Wilcoxon Signed Rank test). When controlling for age, the partial correlations were 0.767 (ST), 0.925 (KEJ) and 0.897 (CV) between the left and right clavicles. In the living, we tested the composite scores of the left and right clavicle against age. In case of one side missing (2 cases by KEJ), the other side was used. For one observer (ST), the mean stage correlated slightly better with age (0.517, p < 0.001, Spearman rank correlation) than the maximum and the minimum stage (0.440 and 0.516, p < 0.001, respectively). At 0.397 (p = 0.003), observer CV showed slightly better correlations with age using the maximum stage, as compared to the mean (0.324) and minimum stage (0.283). At negative values ranging from 0.005 to 0.043,

observer KEJ did not show correlation with age at either the mean, minimum or maximum stage (0.733 < p < 0.974). Using mean stages, the intra-observer Kappa value was 0.662 (n = 47) and 0.526 (n = 53) for autopsy cases and living subjects, respectively, corresponding to substantial and moderate agreement. At prone position, Kappa values rose to 0.675 (n = 32). In autopsy cases, significant Kappa values of moderate agreement were found among observers (K = 0.410–0.447), while only slight agreements (K = 0.077–0.125, p  0.013) were observed in the living. Reliability of stage assessment within and between observers was further tested by comparing the collapsed stages of ‘‘incomplete fusion’’ (stage 3 on both sides) to ‘‘fused’’ (at least one stage 4). As a result, intra-observer reliability improved for autopsy cases as well as living subjects (Table 2). In prone position, intraobserver agreement was substantial (K = 0.788, n = 32). Among autopsy cases, Kappa values rose but remained moderate. In the living, agreement rate improved from slight to moderate between two observers (ST and CV). Between these two observers, 13 subjects disagreed. Of these, 12 were a stage 3 on both sides and the minimum age was 19.8 years. The last case was related to a 24.8 year old male, staged 1 on both sides, scanned in the supine position. In prone position, Kappa values improved among observers to 0.446, n = 32, (ST vs. CV) and 0.040, n = 26 (ST vs. KEJ and KEJ vs. CV). Although observers showed low agreement rate, no subject at or younger than 18 years (4 M, 6 F) were assessed as having ‘‘fused’’ clavicles (Fig. 4)(Table 3). Of these subjects, the highest composite

Table 3 The mean (standard deviation), minimum and maximum age in years of included subjects assessed as ‘‘incomplete fusion’’ ( stage 3 on both sides) and ‘‘fused’’ (at least one stage 4) by the three observers. Observer

ST KEJ CV

Incomplete fusion

Fused

N

Mean + SD (yrs.)

Min–max (yrs.)

N

Mean + SD (yrs.)

Min–max (yrs.)

74 70 67

21.7 (2.8) 22.3 (3.2) 21.5 (2.7)

12.9–31.2 12.9–33.5 12.9–30.7

26 24 33

26.1 (3.1) 24.9 (3.3) 25.5 (3.3)

21.1–33.5 19.8–31.2 19.8–33.5

S. Tangmose et al. / Forensic Science International 234 (2014) 7–12

Fig. 4. Box plot showing the collapsed composite stages of incomplete fusion ( stage 3 on both sides) against fused (at least one stage 4) in autopsy cases (blue dots) and living subjects (red dots). The red line demarks the legal age limit 18 years. When a fused clavicle is assessed on at least one side by all three observers, the chronological age was above 18 years. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

stage assessed was a stage 3 on both sides. Among observers, a stage 3 on both sides was assessed in 0.0% to 60.0% of the subjects. At 21 years (17 M, 6 F), the percentages of subjects assessed as ‘‘fused’’ varied from 9.1% to 17.4% among observers. 4. Discussion A survey performed at our institute showed that nearly all subjects submitted to forensic age estimation exhibit complete skeletal maturity of the hand-wrist bones [16]. In these cases, the Study Group on Forensic Age Diagnostics, AGFAD, recommends a clavicle examination [3]. Our results support that ossification (stage 2) of the medial epiphysis of the clavicle begins at puberty, fusion (stage 3) commences at 16–17 years and generally completes (stage 4) after 19 years of age, in correspondence to the literature [1,17–19]. Although not statistically significant, females tend to mature earlier than males, resembling the sexual differences found in other studies and other parts of the skeleton [19–21]. At wide age-spans, accurate age estimates are difficult to obtain from a single clavicle assessment. However, the clavicle offers important information about age threshold occurring in young adulthood. As indicated by our results, fused clavicles are generally not observed at an age younger than 18 years and exhibited in few at 21 years. When presenting the composite score of stage 3 and 4, the reported probability of an age below 18 years is 6.7–7% in females and 2.6–2.9% in males [19]. The demarcation between stage 2 and 3 may thus be of less consequence, and the use of a collapsed stage 2 and 3 has previously been proposed [19]. As supported by our results, morphology resemblance may risk mistaken a stage 4 for 1 (or opposite) [19]. It has been proposed that this error is minimized when combined with a hand-wrist radiograph [19]. Generally, age estimations performed solely from the clavicle may be difficult. At frequent developmental variants at the sternal end of the clavicle [22–24], observers with radiological experience has been supposed to perform better at assessing clavicles from radiographs [7]. We argue that reliability of stage assessments as well as accuracy of age estimates depend on the level of training in combination to image resolution rather than general radiological expertise.

11

While maturity stages are easily assessed from hand-wrist radiographs and not necessary improve by training [16,25], this does not seem to apply to stages assessed from the clavicle [7]. Our study documents reliability between self-taught observers of different level of training and radiological expertise. As previous documented using CR, our results support that the assessment of the medial epiphyses of the clavicle is neither a simple nor easy method to reproduce by different observers [7]. In the study by Cameriere et al. [7], moderate agreement rates were found using 5 stages by CR. The quality of chest radiographs was suggested to affect agreement rates. By removing movement artefacts and noise, higher image resolution may explain improved reproducibility in autopsy cases. Likewise, reproducibility improved using the prone position. While the intra-observer agreements of our study compare to previous reported substantial Kappa value and Spearman rank correlation of 0.77 and 0.66 (3.0 T MRI) [19], the inter-observer variation is poorer. Previous MR and CT studies have reported inter-observer Kappa values of substantial agreement using a 4-stage system [18,19,21]. However, these studies were primarily based on trained observers whom had received ‘‘extensive image interpretation training’’ or, were based on a single image chosen as the best representative of the stage. A single image does not reflect the typical clinical situation. Although the 3D sequence of our study better visualise the clavicle, it may increase the subjectivity and observer dependence of stage assessments as compared to single images. The disagreements on the composite stages of 3 + 3, 3 + 4 and 4 + 4 may indicate differences in stage assessment when clavicles are nearly fused. Despite the difficulties assessing developmental stages of the clavicles with subsequently low Kappa agreements, no observer assessed a stage 4 on either side in subjects younger than 18 years. However, only 10 cases were younger than 18 years. In MRI, different imaging sequences can be applied to obtain different levels of contrast between structures. Imaging sequences (e.g. T1 or T2 weighted images) have been shown to affect the percentages of assessable images as well as the perceived level of difficulty to apply a stage [19]. Previously, MRI studies using lower field strength for age estimation methods apply T1 weighted imaging sequences [10,26]. We chose a T2 weighted sequence as it discerns the best distinction between the layers of the epiphyseal cartilage [13,27]. Unfortunately, T2 weighted sequences may further reduce image resolution in lower field strength as compared to T1 weighted sequences [28]. Furthermore, T2 weighted images may risk lower agreement rate among observers and a higher tendency for stages to be assessed as advanced as compared to T1 weighted images [29]. At a voxel size at the technical limit compared to the anatomical structure of the clavicular apophysis, image resolution may increase by a further reduction in voxel size. However, this prolongs the examination time and risks the subjects moving. As young individuals undergoing age estimations may originate from ‘‘war zones’’, the risk of shrapnel is apparent. If overlooked, the hazard of metallic implants is reduced at lower field strength [12]. 5. Conclusion Due to the low and inhomogeneous number of cases in this study, conclusions should be drawn carefully. Our study support that clavicle fusion is neither a simple nor easy method to reproduce by different observers. For improved precision, observers need to complete training session and the demarcation between stages must be clearly defined. At poorer resolution, the distinction between developmental stages of the clavicle is difficult. Using the current imaging sequence, the 4-stage system cannot be considered a useful method.

12

S. Tangmose et al. / Forensic Science International 234 (2014) 7–12

Although high accuracy age estimates cannot be obtained, a 2stage system may maintain the ability to exclude an age below 18 years, and thus may serve as an applicable method for age estimation. However, further studies are needed to investigate whether the contrast of assessing ‘‘incomplete fusion’’ from ‘‘fused’’ using a 1.0 T MRI can be used as a method for estimating age in living subjects. If so, lower field strength MRI is a nonionizing method which may serve as a supplement in combination to radiographs of the hand-wrist and teeth without increasing the amount of radiation exposure. Further, low field MRI serves well as a supplement to CT imaging prior to medico-legal autopsy. Acknowledgments The authors would like to thank Professor, Dr. Med. Carsten Thomsen and Radiographer Joan Balslev Andersen at Rigshospitalet for assisting with the protocol as well as Radiographers Sabeen Arshad and Bo Rasmussen and Forensic Technicians at the Section of Forensic Pathology, University of Copenhagen for implementing the data collection during the daily routine. Finally, thanks to all the young individuals whom voluntarily received an MRI scan for this study. References [1] S. Black, L. Scheuer, Age changes in the clavicle: from the early neonatal period to skeletal maturity, Int. J. Osteoarchaeol. 6 (1996) 425–434. [2] E. Hilleweg, J. De Tobel, O. Cuche, P. Vandemaele, M. Piette, K. Verstraete, Magnetic resonance imaging of the medial extremity of the clavicle in forensic bone age determination: a new four-minute approach, Eur. Radiol. 21 (2011) 757–767. [3] A. Schmeling, C. Grundmann, A. Fuhrmann, H.J. Kaatsch, B. Knell, F. Ramsthaler, et al., Criteria for age estimation in living individuals, Int. J. Leg. Med. 122 (2008) 457–460. [4] M. Kellinghaus, R. Schulz, V. Vieth, S. Schmidt, H. Pfeiffer, A. Schmeling, Enhanced possibilities to make statements on the ossification status of the medial clavicular epiphysis using an amplified staging scheme in evaluating thin-slice CT scans, Int. J. Leg. Med. 124 (2010) 321–325. [5] A. Schmeling, R. Schulz, W. Reisinger, M. Muhler, K.D. Wernecke, G. Geserick, Studies on the time frame for ossification of the medial clavicular epiphyseal cartilage in conventional radiography, Int. J. Leg. Med. 118 (2004) 5–8. [6] F. Ramsthaler, P. Proschek, W. Betz, M.A. Verhoff, How reliable are the risk estimates for X-ray examinations in forensic age estimations? A safety update, Int. J. Leg. Med. 123 (2009) 199–204. [7] R. Cameriere, S. De Luca, D. Angelis, V. Merelli, A. Giuliodori, M. Cingolani, et al., Reliability of schmeling’ stages of ossification of medial clavicular epiphyses and its validity to assess 18 years of age in living subjects, Int. J. Leg. Med. 126 (2012) 923–932. [8] R. Schulz, M. Muhler, W. Reisinger, W. Schmidt, A. Schmeling, Radiographic staging of ossification of the medial clavicluar epiphysis, Int. J. Leg. Med. 122 (2008) 55–58.

[9] M.S. Pearce, J.A. Salotti, M.P. Little, K. McHugh, C. Lee, K.P. Kim, et al., Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study, Lancet 380 (2012) 499–505. [10] S. Schmidt, M. Muhler, A. Schmeling, W. Reisinger, R. Schulz, Magnetic resonance imaging of clavicular ossification, Int. J. Leg. Med. 121 (2007) 321–324. [11] M. Baclivo, S. Winklhofer, G. Hatch, G. Ampanosi, M. Thali, T. Ruder, The rise of forensic and post-mortem radiology—analysis of the literature between the year 2000 and 2011, J. Forensic Radiol. Imaging 1 (2013) 3–9. [12] B.J. Soher, B.M. Dale, E.M. Merkle, A review of MR physics: 3 T versus 1.5 T, Magn. Reson. Imaging Clin. N. Am. 15 (2007) 277–279. [13] P.C. Khanna, M.M. Thapa, The growing skeleton: MR imaging appearances of developing cartilage, Magn. Reson. Imaging Clin. N. Am. 17 (3) (2009) 411–421. [14] M. Muhler, R. Schulz, W. Schmidt, A. Schmeling, W. Reisinger, The influence on slice thickness on assessment of clavicle ossification in forensic age diagnostics, Int. J. Leg. Med. 120 (2006) 15–17. [15] J.R. Landis, G.G. Koch, The measurement of observer agreement for categorical data, Biometrics 33 (1977) 159–174. [16] N. Lynnerup, E. Belard, K. Buch-Olsen, B. Sejrsen, K. Damgaard-Pedersen, Intraand interobserver error of the Greulich–Pyle method as used on a Danish forensic sample, Forensic Sci. Int. 179 (2008) 242–246. [17] R. Schulz, M. Muhler, S. Mutze, S. Schmidt, W. Reisinger, A. Schmeling, Studies on the time frame for ossification of the medial epiphysis of the clavicle as revealed by CT scans, Int. J. Leg. Med. 119 (2005) 142–145. [18] D. Schulze, U. Rother, A. Fuhrmann, S. Richel, G. Faulmann, M. Heiland, Correlation of age and ossification of the medial clavicular epiphysis using computed tomography, Forensic Sci. Int. 158 (2006) 184–189. [19] E. Hilleweg, J. Degroote, T. Van der Taelt, A. Visscher, P. Vandemaele, B. Lutin, et al., Magnetic resonance imaging of the sternal extremity of the clavicle in forensic age estimation: towards more sound age estimates, Int. J. Leg. Med. (2012). [20] L. Meijerman, G.J. Maat, R. Schulz, A. Schmeling, Variables affecting the probability of complete fusion of the medial clavicular epiphysis, Int. J. Leg. Med. 121 (2007) 463–468. [21] R.B. Bassed, O.H. Drummer, C. Briggs, A. Valenzuela, Age estimation and the medial clavicular epiphysis: analysis of the age of majority in an Australian population using computed tomography, Forensic Sci. Med. Pathol. 7 (2011) 148–154. [22] A.G. Jurik, Imaging of the Sternocostoclavicular Region, Springer-Verlag Berlin Heidelberg, Berlin, 2007. [23] R. Kumar, J.E. Madewell, L.E. Swischuk, M.M. Lindell, R. David, The clavicle: normal and abnormal, Radiographics 9 (1998) 677–706. [24] M. Kellinghaus, R. Schulz, V. Vieth, S. Schmidt, A. Schmeling, 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. Leg. Med. 124 (2010) 149–154. [25] A.F. Roche, G.H. Davila, B.A. Pasternack, M.J. Walton, Some factors influencing the replicability of assessments of skeletal maturity (Greulich–Pyle), Am. J. Roentgenol. Radium Ther. Nucl. Med. 109 (1970) 299–306. [26] J. Dvorak, J. George, A. Junge, J. Hoddler, Age determination by magnetic resonance imaging of the wrist in adolescent male football players, Br. J. Sports Med. 41 (2007) 45–52. [27] D. Jaramillo, S.A. Connolly, R.V. Mulkern, F. Shapiro, Developing epiphysis: MR imaging characteristics and histologic correlation in the newborn lamb, Radiology 2007 (1998) 637–645. [28] B.K. Rutt, D.H. Lee, The impact of field strength on image quality in MRI, J. Magn. Reson. Imaging 6 (1996) 57–62. [29] Y. Shimada, D. Shimao, J. Kobayashi, C. Nakahori, M. Shimada, H. Fujimoto, et al., Comparison of MR images for age determination; T1 weighted images (T1WI) versus T2* weighted images (T2*WI), Asian J. Sports Med. 3 (2012) 47–52.

Forensic age estimation from the clavicle using 1.0T MRI--preliminary results.

As forensic age estimations in the living are performed without medical indication, there is a need for the development of non-ionizing methods. This ...
880KB Sizes 1 Downloads 0 Views