Pub Date : 2025-04-25DOI: 10.1016/j.fri.2025.200635
Jamie Elifritz , Heather Jarrell , Fabrice Dedouit , Laura Filograna , ISFRI guidelines Working Group
Postmortem computed tomography (PMCT) has emerged as a valuable tool in forensic investigations, supporting the investigation of suspected overdoses. While not specific, Cerebral edema, pulmonary edema, and urinary bladder distention (the overdose triad) can suggest overdose in unsuspected cases. Furthermore, a high-density basal layer in the stomach may indicate intentional therapeutic medical overdose. Challenges include short agonal intervals and decomposition changes. Confirmatory blood toxicology is necessary. Dual-energy computed tomography (DECT) can play a role in differentiation of material contributing to dense basal layers and body packing scenarios. PMCT serves as a valuable complement to autopsy findings, aiding in the assessment of internal pathology while also offering a non-invasive alternative in specific forensic contexts where autopsy may not be performed.
{"title":"ISFRI Guidelines Working Group: Best practice standards for non-contrast postmortem computed tomography (PMCT) of overdose","authors":"Jamie Elifritz , Heather Jarrell , Fabrice Dedouit , Laura Filograna , ISFRI guidelines Working Group","doi":"10.1016/j.fri.2025.200635","DOIUrl":"10.1016/j.fri.2025.200635","url":null,"abstract":"<div><div>Postmortem computed tomography (PMCT) has emerged as a valuable tool in forensic investigations, supporting the investigation of suspected overdoses. While not specific, Cerebral edema, pulmonary edema, and urinary bladder distention (the overdose triad) can suggest overdose in unsuspected cases. Furthermore, a high-density basal layer in the stomach may indicate intentional therapeutic medical overdose. Challenges include short agonal intervals and decomposition changes. Confirmatory blood toxicology is necessary. Dual-energy computed tomography (DECT) can play a role in differentiation of material contributing to dense basal layers and body packing scenarios. PMCT serves as a valuable complement to autopsy findings, aiding in the assessment of internal pathology while also offering a non-invasive alternative in specific forensic contexts where autopsy may not be performed.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"42 ","pages":"Article 200635"},"PeriodicalIF":0.8,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144535010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Identifying the cause of death using postmortem CT images is crucial since it provides a non-invasive, objective approach for forensic investigations while offering significant advantages in terms of time efficiency and cost-effectiveness compared to traditional autopsy methods. However, due to varied lung conditions in the postmortem CT images, a standardized method to diagnose drowning using CT images has not been established. This study aimed to devise a deep-learning-aided framework for diagnosing drowning from postmortem lung CT images. First, to find the suitable convolutional neural network (CNN) architecture for classifying lung CT images into drowning and non-drowning cases, three well-known CNNs, AlexNet, VGG16, and MobileNet, were trained with a single-institute postmortem CT image dataset and the performance and generalizability were also evaluated using images extracted from a public decedent CT image database. The results showed that VGG16 architecture outperformed the three models with the highest mean AUC-ROC and accuracy values of 88.42 % and 80.56 % respectively for drowning image classification, as well as the highest generalizability with an AUC-ROC of 71.79 % on a public image dataset. Additionally, the case-based diagnosis was performed using probability scores given from the model to each slice taken in the same subject. The final diagnosis accuracy was 96 % on the original dataset and 79 % on the public dataset, showing the strong potential that the devised framework can be used as a screening tool to identify drowning cases using postmortem CT images.
{"title":"A deep-learning-aided diagnosis of drowning using post-mortem lung computed tomography","authors":"Amber Habib Qureshi, Takuro Ishii, Yoshifumi Saijo","doi":"10.1016/j.fri.2025.200629","DOIUrl":"10.1016/j.fri.2025.200629","url":null,"abstract":"<div><div>Identifying the cause of death using postmortem CT images is crucial since it provides a non-invasive, objective approach for forensic investigations while offering significant advantages in terms of time efficiency and cost-effectiveness compared to traditional autopsy methods. However, due to varied lung conditions in the postmortem CT images, a standardized method to diagnose drowning using CT images has not been established. This study aimed to devise a deep-learning-aided framework for diagnosing drowning from postmortem lung CT images. First, to find the suitable convolutional neural network (CNN) architecture for classifying lung CT images into drowning and non-drowning cases, three well-known CNNs, AlexNet, VGG16, and MobileNet, were trained with a single-institute postmortem CT image dataset and the performance and generalizability were also evaluated using images extracted from a public decedent CT image database. The results showed that VGG16 architecture outperformed the three models with the highest mean AUC-ROC and accuracy values of 88.42 % and 80.56 % respectively for drowning image classification, as well as the highest generalizability with an AUC-ROC of 71.79 % on a public image dataset. Additionally, the case-based diagnosis was performed using probability scores given from the model to each slice taken in the same subject. The final diagnosis accuracy was 96 % on the original dataset and 79 % on the public dataset, showing the strong potential that the devised framework can be used as a screening tool to identify drowning cases using postmortem CT images.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"41 ","pages":"Article 200629"},"PeriodicalIF":0.8,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143886972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We aimed to evaluate the correlation between stature and vertical skull measurements based on three-dimensional (3D) computed tomography (CT) images, develop a stature estimation formula, and validate it in a Japanese population. The “training” and “validation” datasets consisted of 275 and 49 identified individuals who underwent postmortem CT. Two skull measurements, the linear distances from the basion to the bregma (Basion–Bregma) and the posterior nasal spine to the bregma (PNS–Bregma), were obtained from 3D CT images that solely extracted cranial data. Pearson product-moment correlation coefficients assessed stature-skull correlations. Multiple regression analysis was performed to assess whether stature was dependent on sex. A stature estimation formula was developed based on the regression analysis. Validation tests were performed for each formula. Significant correlations were observed between stature and skull measurements. The correlation coefficients were 0.790 for stature and Basion–Bregma, and 0.782 for stature and PNS–Bregma. Sex status was statistically significant as an independent variable in regression analysis and influences the estimation of stature. For the stature estimation formula, the coefficient of determination adjusted for the degree of freedom (R*2) was 0.730, and the standard error of estimation (SEE) was 5.55 cm when using three variables: sex status, Basion–Bregma, and PNS–Bregma. The validation test exhibited high reliability; mean error (MA) = 0.10 cm and mean absolute error (MAE) = 4.64 cm. Vertical skull measurements from 3D CT images may be helpful for stature estimation in Japanese individuals, particularly in cases where better predictors are unavailable.
{"title":"Stature estimation based on vertical measurements of the skull using three-dimensional images from postmortem computed tomography in a Japanese population","authors":"Shoken Suzuki , Maki Ohtani , Yuhei Matsuo , Masayuki Fukuda , Sohtaro Mimasaka","doi":"10.1016/j.fri.2025.200628","DOIUrl":"10.1016/j.fri.2025.200628","url":null,"abstract":"<div><div>We aimed to evaluate the correlation between stature and vertical skull measurements based on three-dimensional (3D) computed tomography (CT) images, develop a stature estimation formula, and validate it in a Japanese population. The “training” and “validation” datasets consisted of 275 and 49 identified individuals who underwent postmortem CT. Two skull measurements, the linear distances from the basion to the bregma (Basion–Bregma) and the posterior nasal spine to the bregma (PNS–Bregma), were obtained from 3D CT images that solely extracted cranial data. Pearson product-moment correlation coefficients assessed stature-skull correlations. Multiple regression analysis was performed to assess whether stature was dependent on sex. A stature estimation formula was developed based on the regression analysis. Validation tests were performed for each formula. Significant correlations were observed between stature and skull measurements. The correlation coefficients were 0.790 for stature and Basion–Bregma, and 0.782 for stature and PNS–Bregma. Sex status was statistically significant as an independent variable in regression analysis and influences the estimation of stature. For the stature estimation formula, the coefficient of determination adjusted for the degree of freedom (<em>R</em>*<sup>2</sup>) was 0.730, and the standard error of estimation (SEE) was 5.55 cm when using three variables: sex status, Basion–Bregma, and PNS–Bregma. The validation test exhibited high reliability; mean error (MA) = 0.10 cm and mean absolute error (MAE) = 4.64 cm. Vertical skull measurements from 3D CT images may be helpful for stature estimation in Japanese individuals, particularly in cases where better predictors are unavailable.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"41 ","pages":"Article 200628"},"PeriodicalIF":0.8,"publicationDate":"2025-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868382","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Assessing skeletal maturity using epiphyseal and morphological features with modern, reliable evaluation protocols is crucial for human identification efforts and paediatric growth monitoring. This study aims to develop and validate a scoring system for knee skeletal development on post-mortem computed tomography (PMCT) and magnetic resonance imaging (MRI) acquired from Australian and New Mexican children.
Materials & Methods
A protocol for the skeletal knee was developed on 30 PMCT and 30 T2-weighted MRI scans of subadults aged eight- to- 22 years. DICOM image stacks from a Brisbane children’s hospital and the New Mexico Decedent Image Database (NMDID) underwent multiplanar reconstruction and anatomical alignment. The protocol comprised a three- to- six stage scoring process at four epiphyseal fusion and seven maturity indicator sites. Three observers of varying experience levels assessed the images across three days, with reliability quantified using an intraclass correlation coefficient (ICC).
Results
The protocol demonstrated high reliability and consistency, with excellent intraobserver agreement for CT (ICC = 0.985 (95 % CI: 0.93-1.00)) and MRI (ICC = 0.979 (95 % CI: 0.85-1.00)). Mean inter-observer reliability measures were good for CT (ICC = 0.886 (95 % CI: 0.75-0.95)) and MRI (ICC = 0.852 (95 % CI: 0.68-0.95)). The tibial tubercle demonstrated the most variability and long-bone epiphyseal union the least
Conclusions
This research presents a highly reproducible method for assessing skeletal development of the knee in subadults, aligned with modern imaging standards. The methodology will have application in forensic human identification, age confirmation and clinical growth assessment
{"title":"Assessment of knee ossification timings: Development and validation of an ordinal scoring protocol for age estimation using medical imaging","authors":"Taliah Swart , Samantha K. Rowbotham , Soren Blau , Nicolene Lottering","doi":"10.1016/j.fri.2025.200625","DOIUrl":"10.1016/j.fri.2025.200625","url":null,"abstract":"<div><h3>Objectives</h3><div>Assessing skeletal maturity using epiphyseal and morphological features with modern, reliable evaluation protocols is crucial for human identification efforts and paediatric growth monitoring. This study aims to develop and validate a scoring system for knee skeletal development on post-mortem computed tomography (PMCT) and magnetic resonance imaging (MRI) acquired from Australian and New Mexican children.</div></div><div><h3>Materials & Methods</h3><div>A protocol for the skeletal knee was developed on 30 PMCT and 30 T2-weighted MRI scans of subadults aged eight- to- 22 years. DICOM image stacks from a Brisbane children’s hospital and the New Mexico Decedent Image Database (NMDID) underwent multiplanar reconstruction and anatomical alignment. The protocol comprised a three- to- six stage scoring process at four epiphyseal fusion and seven maturity indicator sites. Three observers of varying experience levels assessed the images across three days, with reliability quantified using an intraclass correlation coefficient (ICC).</div></div><div><h3>Results</h3><div>The protocol demonstrated high reliability and consistency, with excellent intraobserver agreement for CT (ICC = 0.985 (95 % CI: 0.93-1.00)) and MRI (ICC = 0.979 (95 % CI: 0.85-1.00)). Mean inter-observer reliability measures were good for CT (ICC = 0.886 (95 % CI: 0.75-0.95)) and MRI (ICC = 0.852 (95 % CI: 0.68-0.95)). The tibial tubercle demonstrated the most variability and long-bone epiphyseal union the least</div></div><div><h3>Conclusions</h3><div>This research presents a highly reproducible method for assessing skeletal development of the knee in subadults, aligned with modern imaging standards. The methodology will have application in forensic human identification, age confirmation and clinical growth assessment</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"41 ","pages":"Article 200625"},"PeriodicalIF":0.8,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Estimating hemothorax volume via postmortem computed tomography (PMCT) remains challenging because of postmortem artifacts that can impact interpretation and accuracy. This study aimed to evaluate the accuracy of PMCT in estimating hemothorax volume compared with standard autopsy.
Methods
Forty deceased individuals who underwent both PMCT and autopsy were examined. PMCT volumes were manually segmented, and the results were compared with autopsy findings. Spearman's rank correlation and paired t-tests were used to assess accuracy.
Results
PMCT showed a high diagnostic accuracy for hemothorax, with correlation coefficients of 0.859 and 0.794 on the left and right sides, respectively. However, the mean absolute percentage error (MAPE) for volume estimation was relatively high, suggesting caution when relying solely on PMCT for volume estimation.
Conclusion
PMCT is a reliable tool for diagnosing hemothorax; however, its accuracy in volume estimation remains limited for manual segmentation methods. Further refinement of the imaging techniques is required for more precise volume measurements.
{"title":"High diagnostic accuracy of postmortem CT for hemothorax with volume estimation challenges: A comparative study with autopsy","authors":"Punpramepree Yeesakhorn , Wanlapha Tungsub , Nitima Saksobhavivat , Wisarn Worasuwannarak","doi":"10.1016/j.fri.2025.200626","DOIUrl":"10.1016/j.fri.2025.200626","url":null,"abstract":"<div><h3>Background</h3><div>Estimating hemothorax volume via postmortem computed tomography (PMCT) remains challenging because of postmortem artifacts that can impact interpretation and accuracy. This study aimed to evaluate the accuracy of PMCT in estimating hemothorax volume compared with standard autopsy.</div></div><div><h3>Methods</h3><div>Forty deceased individuals who underwent both PMCT and autopsy were examined. PMCT volumes were manually segmented, and the results were compared with autopsy findings. Spearman's rank correlation and paired t-tests were used to assess accuracy.</div></div><div><h3>Results</h3><div>PMCT showed a high diagnostic accuracy for hemothorax, with correlation coefficients of 0.859 and 0.794 on the left and right sides, respectively. However, the mean absolute percentage error (MAPE) for volume estimation was relatively high, suggesting caution when relying solely on PMCT for volume estimation.</div></div><div><h3>Conclusion</h3><div>PMCT is a reliable tool for diagnosing hemothorax; however, its accuracy in volume estimation remains limited for manual segmentation methods. Further refinement of the imaging techniques is required for more precise volume measurements.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"41 ","pages":"Article 200626"},"PeriodicalIF":0.8,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143855481","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-17DOI: 10.1016/j.fri.2025.200624
Deepa Jatti Patil, Chandramani B. More, Rashmi Venkatesh
The study aimed to determine age by assessing the discernibility of the periodontal ligament (PDL) and root canal (RC) on panoramic radiographs of mandibular third molars. In this retrospective study, 2000 panoramic radiographs of individuals aged between 16 to 40 were analysed, including both males and females. The radiographic discernibility of PDL and RC in mandibular third molars was assessed according to the study by Olze et al. which was categorised into four stages. At each stage, the minimum, maximum, and standard deviation were assessed. Statistical analysis was conducted to examine the relationship between age, sex, and PDL/RC stage. There was a notable disparity in the average age of individuals at different stages of PDL and RC. There was a considerable increase in the average age from PDL & RC stage 0 to stage 3. By considering the minimum and maximum values for each stage, individuals can be classified as being older than 17 years if they are in stage 1, and older than 20 years if they are in stages 2 and 3. These classifications are determined based on the combined results of the PDL and RC stages. The radiographic discernibility of PDL and RC can be utilised as a promising method to determine age in the western Indian population.
{"title":"Dental age estimation based on imaging of lower third molars in Western Indian population","authors":"Deepa Jatti Patil, Chandramani B. More, Rashmi Venkatesh","doi":"10.1016/j.fri.2025.200624","DOIUrl":"10.1016/j.fri.2025.200624","url":null,"abstract":"<div><div>The study aimed to determine age by assessing the discernibility of the periodontal ligament (PDL) and root canal (RC) on panoramic radiographs of mandibular third molars. In this retrospective study, 2000 panoramic radiographs of individuals aged between 16 to 40 were analysed, including both males and females. The radiographic discernibility of PDL and RC in mandibular third molars was assessed according to the study by Olze et al. which was categorised into four stages. At each stage, the minimum, maximum, and standard deviation were assessed. Statistical analysis was conducted to examine the relationship between age, sex, and PDL/RC stage. There was a notable disparity in the average age of individuals at different stages of PDL and RC. There was a considerable increase in the average age from PDL & RC stage 0 to stage 3. By considering the minimum and maximum values for each stage, individuals can be classified as being older than 17 years if they are in stage 1, and older than 20 years if they are in stages 2 and 3. These classifications are determined based on the combined results of the PDL and RC stages. The radiographic discernibility of PDL and RC can be utilised as a promising method to determine age in the western Indian population.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"41 ","pages":"Article 200624"},"PeriodicalIF":0.8,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143879410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-17DOI: 10.1016/j.fri.2025.200627
Kirthika Ravi , Siddhartha Das , Ambika Prasad Patra , Deepak Barathi Subramania , Harichandrakumar Kottyen Thazhath
Background
Stature estimation contributes to the identification of an individual which is one of the objectives of a medicolegal autopsy. Stature can be estimated by measuring various landmarks of the cranium. Owing to the geographical variations, the regression formula used for one population may not be applicable to other populations. This CT scan study was conducted with an aim to develop regression formulas for the different cranial parameters in a South Indian adult population.
Methodology
511 patients scheduled for elective CT scans of the head and neck were recruited. Twenty-nine cranial variables were studied in each of these patients. Simple and multivariate linear regression was performed to establish a predictive stature estimation model. Pearson correlation and the predictive stature estimation model were considered significant if the P value was ≤ 0.05.
Results
All the cranial measurements showed a statistically significant correlation with stature in the overall population except for right orbital height, left orbital height and minimum distance between the condyles. The proportion of variance of stature explained by the model was found to be 27 % for the overall population, whereas it was 20 % and 21 % respectively for the males and females.
Conclusion
Our results suggest that the studied cranial measurements have a positive correlation with stature and can be used to estimate the stature, but the R2 values are not so encouraging.
{"title":"Stature estimation and craniometry–a computed tomography scan based study in South Indian adult population","authors":"Kirthika Ravi , Siddhartha Das , Ambika Prasad Patra , Deepak Barathi Subramania , Harichandrakumar Kottyen Thazhath","doi":"10.1016/j.fri.2025.200627","DOIUrl":"10.1016/j.fri.2025.200627","url":null,"abstract":"<div><h3>Background</h3><div>Stature estimation contributes to the identification of an individual which is one of the objectives of a medicolegal autopsy. Stature can be estimated by measuring various landmarks of the cranium. Owing to the geographical variations, the regression formula used for one population may not be applicable to other populations. This CT scan study was conducted with an aim to develop regression formulas for the different cranial parameters in a South Indian adult population.</div></div><div><h3>Methodology</h3><div>511 patients scheduled for elective CT scans of the head and neck were recruited. Twenty-nine cranial variables were studied in each of these patients. Simple and multivariate linear regression was performed to establish a predictive stature estimation model. Pearson correlation and the predictive stature estimation model were considered significant if the <em>P</em> value was ≤ 0.05.</div></div><div><h3>Results</h3><div>All the cranial measurements showed a statistically significant correlation with stature in the overall population except for right orbital height, left orbital height and minimum distance between the condyles. The proportion of variance of stature explained by the model was found to be 27 % for the overall population, whereas it was 20 % and 21 % respectively for the males and females.</div></div><div><h3>Conclusion</h3><div>Our results suggest that the studied cranial measurements have a positive correlation with stature and can be used to estimate the stature, but the R<sup>2</sup> values are not so encouraging.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"41 ","pages":"Article 200627"},"PeriodicalIF":0.8,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143865102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-28DOI: 10.1016/j.fri.2024.200622
Harry Perkins , Thao Liang Chiam , Alex Forrest , Denice Higgins
Background
Forensic odontology is crucial for human identification, especially in disaster scenarios, using comparisons between antemortem and postmortem dental data. Advances in 3D imaging have shifted practices from traditional 2D methods to 3D superimposition techniques, necessitating a comprehensive review. This scoping review maps current 3D superimposition methodologies in forensic odontology, focusing on key processes, sources of error, and research gaps.
Methods
We used Arksey and O'Malley's framework, searching PubMed, Embase, Scopus, and DOSS for studies from January 2017. Search strategies incorporated MeSH and Emtree terms, Boolean operators, and truncations. Inclusion criteria required studies to utilize 3D superimposition techniques for comparing dental imaging, with exclusions for 2D imaging, non-forensic focus, and inaccessible texts. Data were extracted on anatomical features, imaging techniques, methods, and outcomes.
Results
From 545 records, 20 studies met inclusion criteria. Most employed surface-based superimposition. Methodologies varied widely, with inconsistent software use and a lack of standardization. Root Mean Square (RMS) values were commonly used to assess alignment, but thresholds differed significantly across studies. Key challenges include operator variability and limited access to affordable software.
Conclusions
The rapid advancement of 3D imaging in forensic odontology highlights the need for standardized methods. While surface-based techniques are promising, establishing uniform benchmarks and developing open-source tools are crucial for improving reliability and global adoption.
{"title":"3D dental images in forensic odontology: A scoping review of superimposition approaches utilizing 3D imaging","authors":"Harry Perkins , Thao Liang Chiam , Alex Forrest , Denice Higgins","doi":"10.1016/j.fri.2024.200622","DOIUrl":"10.1016/j.fri.2024.200622","url":null,"abstract":"<div><h3>Background</h3><div>Forensic odontology is crucial for human identification, especially in disaster scenarios, using comparisons between antemortem and postmortem dental data. Advances in 3D imaging have shifted practices from traditional 2D methods to 3D superimposition techniques, necessitating a comprehensive review. This scoping review maps current 3D superimposition methodologies in forensic odontology, focusing on key processes, sources of error, and research gaps.</div></div><div><h3>Methods</h3><div>We used Arksey and O'Malley's framework, searching PubMed, Embase, Scopus, and DOSS for studies from January 2017. Search strategies incorporated MeSH and Emtree terms, Boolean operators, and truncations. Inclusion criteria required studies to utilize 3D superimposition techniques for comparing dental imaging, with exclusions for 2D imaging, non-forensic focus, and inaccessible texts. Data were extracted on anatomical features, imaging techniques, methods, and outcomes.</div></div><div><h3>Results</h3><div>From 545 records, 20 studies met inclusion criteria. Most employed surface-based superimposition. Methodologies varied widely, with inconsistent software use and a lack of standardization. Root Mean Square (RMS) values were commonly used to assess alignment, but thresholds differed significantly across studies. Key challenges include operator variability and limited access to affordable software.</div></div><div><h3>Conclusions</h3><div>The rapid advancement of 3D imaging in forensic odontology highlights the need for standardized methods. While surface-based techniques are promising, establishing uniform benchmarks and developing open-source tools are crucial for improving reliability and global adoption.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"40 ","pages":"Article 200622"},"PeriodicalIF":0.8,"publicationDate":"2024-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143132013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-28DOI: 10.1016/j.fri.2024.200621
Jiwon Choi , Seongwon Choi , Arthur Porto , Harry Perkins , Sanmarié Schlebusch , Denice Higgins , Ove A. Peters , Christine I. Peters , Alex Forrest
Three-dimensional (3D) imaging techniques including radiographic and optical surface scans are used in many medical fields including dentistry. Comparison of the resulting antemortem (AM) and postmortem (PM) data has been limited, partly due to the absence of specialised, affordable software. To address this gap, we introduce ModelMatch3D, an open-source software built upon the established 3D Slicer platform, designed to automate general 3D comparison tasks and enable rapid comparison of 3D surface data. It requires minimal preparation of datasets prior to use.
ModelMatch3D was evaluated using de-identified 3D dental scans sourced from a collection at the University of Adelaide. Dental data was selected due to its recognised importance as one of the three major identifiers by INTERPOL. Although we tested it with dental data, it can be applied to any 3D surface datasets of distinctive shape, extending its utility beyond forensic science.
ModelMatch3D runs on all major operating systems ensuring wide accessibility. It runs well on lower-end systems without discrete graphics capability. Its broad compatibility and minimal computing requirements enable its deployment in diverse forensic environments, providing an easy-to-use, automated method for rapid comparison and where multiple victims are involved, ranking of matches for subsequent expert inspection.
Technically, ModelMatch3D employs advanced algorithms for point cloud processing and feature extraction which enable rapid handling of substantial databases with no need for preprocessing.
We believe that ModelMatch3D fills a gap in forensic identification and opens avenues for comparison of other hard tissues and structures, providing a robust platform for enhancing computational forensics.
{"title":"Development of automatic 3D model comparison (ModelMatch3D) for forensic identification and testing using odontology data","authors":"Jiwon Choi , Seongwon Choi , Arthur Porto , Harry Perkins , Sanmarié Schlebusch , Denice Higgins , Ove A. Peters , Christine I. Peters , Alex Forrest","doi":"10.1016/j.fri.2024.200621","DOIUrl":"10.1016/j.fri.2024.200621","url":null,"abstract":"<div><div>Three-dimensional (3D) imaging techniques including radiographic and optical surface scans are used in many medical fields including dentistry. Comparison of the resulting antemortem (AM) and postmortem (PM) data has been limited, partly due to the absence of specialised, affordable software. To address this gap, we introduce ModelMatch3D, an open-source software built upon the established 3D Slicer platform, designed to automate general 3D comparison tasks and enable rapid comparison of 3D surface data. It requires minimal preparation of datasets prior to use.</div><div>ModelMatch3D was evaluated using de-identified 3D dental scans sourced from a collection at the University of Adelaide. Dental data was selected due to its recognised importance as one of the three major identifiers by INTERPOL. Although we tested it with dental data, it can be applied to any 3D surface datasets of distinctive shape, extending its utility beyond forensic science.</div><div>ModelMatch3D runs on all major operating systems ensuring wide accessibility. It runs well on lower-end systems without discrete graphics capability. Its broad compatibility and minimal computing requirements enable its deployment in diverse forensic environments, providing an easy-to-use, automated method for rapid comparison and where multiple victims are involved, ranking of matches for subsequent expert inspection.</div><div>Technically, ModelMatch3D employs advanced algorithms for point cloud processing and feature extraction which enable rapid handling of substantial databases with no need for preprocessing.</div><div>We believe that ModelMatch3D fills a gap in forensic identification and opens avenues for comparison of other hard tissues and structures, providing a robust platform for enhancing computational forensics.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"40 ","pages":"Article 200621"},"PeriodicalIF":0.8,"publicationDate":"2024-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143131895","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-23DOI: 10.1016/j.fri.2024.200620
Jessika Camatti , Anna Laura Santunione , Stefano Draisci , Drago Antonella , Maria Grazia Amorico , Guido Ligabue , Enrico Silingardi , Pietro Torricelli , Rossana Cecchi
Epicardial fat volume (EFV) has recently been recognized as a good predictor of multivessel coronary artery disease, and the Coronary Artery Calcium Score (CACS) is a parameter that estimates the amount of calcium in the coronary tree. Both of these parameters can be assessed non-invasively by cardiac computed tomography. Previous studies have investigated a correlation between autopsy results and radiologically calculated EFV and CACS.
The present study aims to investigate a correlation between EFV and other radiological (CACS and the presence of coronary artery opacification defects on Multi-Phase Post-Mortem Computed Tomography Angiography (MPMCTA)) and autoptic (presence of coronary stenosis) findings, in order to verify whether EFV can be considered a good predictor of radiological and autoptic coronary findings.
A cohort of 21 subjects who died suddenly was examined. Firstly, MPMCTA was performed, then autopsy was carried out. EFV and CACS were radiologically calculated, the detection of opacification defects on MPMCTA was investigated and the presence of coronary stenoses on autopsy was assessed.
21 deceased individuals (51 ± 10,77 years; 19 men) were evaluated. Statistically significant correlations were found between levels of EFV > 125 mL (cut-off indicated for prognostic risk stratification) and CACS > 0 (signifying the presence of coronary calcifications), opacification defects on MPMCTA, and coronary stenosis on autopsy.
The volume of the epicardial fat, detected radiologically, is a promising additional tool in the assessment and risk stratification for sudden death. Further research is needed to better explore the application of radiologically calculated EFV in cases of sudden death.
{"title":"Correlation between epicardial fat volume and postmortem radiological and autopsy findings in cases of sudden death: A pilot study","authors":"Jessika Camatti , Anna Laura Santunione , Stefano Draisci , Drago Antonella , Maria Grazia Amorico , Guido Ligabue , Enrico Silingardi , Pietro Torricelli , Rossana Cecchi","doi":"10.1016/j.fri.2024.200620","DOIUrl":"10.1016/j.fri.2024.200620","url":null,"abstract":"<div><div>Epicardial fat volume (EFV) has recently been recognized as a good predictor of multivessel coronary artery disease, and the Coronary Artery Calcium Score (CACS) is a parameter that estimates the amount of calcium in the coronary tree. Both of these parameters can be assessed non-invasively by cardiac computed tomography. Previous studies have investigated a correlation between autopsy results and radiologically calculated EFV and CACS.</div><div>The present study aims to investigate a correlation between EFV and other radiological (CACS and the presence of coronary artery opacification defects on Multi-Phase Post-Mortem Computed Tomography Angiography (MPMCTA)) and autoptic (presence of coronary stenosis) findings, in order to verify whether EFV can be considered a good predictor of radiological and autoptic coronary findings.</div><div>A cohort of 21 subjects who died suddenly was examined. Firstly, MPMCTA was performed, then autopsy was carried out. EFV and CACS were radiologically calculated, the detection of opacification defects on MPMCTA was investigated and the presence of coronary stenoses on autopsy was assessed.</div><div>21 deceased individuals (51 ± 10,77 years; 19 men) were evaluated. Statistically significant correlations were found between levels of EFV > 125 mL (cut-off indicated for prognostic risk stratification) and CACS > 0 (signifying the presence of coronary calcifications), opacification defects on MPMCTA, and coronary stenosis on autopsy.</div><div>The volume of the epicardial fat, detected radiologically, is a promising additional tool in the assessment and risk stratification for sudden death. Further research is needed to better explore the application of radiologically calculated EFV in cases of sudden death.</div></div>","PeriodicalId":40763,"journal":{"name":"Forensic Imaging","volume":"40 ","pages":"Article 200620"},"PeriodicalIF":0.8,"publicationDate":"2024-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143132014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号