Artificial intelligence-based forensic sex determination of East Asian cadavers from skull morphology.

Identification of unknown cadavers is an important task for forensic scientists. Forensic scientists attempt to identify skeletal remains based on factors including age, sex, and dental treatment remains. Forensic scientists commonly consider skull or pelvic shape to evaluate the sex; however, these evaluations require sufficient experience and knowledge and lack objectivity and reproducibility. To ensure objectivity and reproducibility for sex evaluation, we applied a gated attention-based multiple-instance learning model to three-dimensional (3D) skull images reconstructed from postmortem head computed tomography scans. We preprocessed the images, trained with 864 training data, validated the model with 124 validation data, and evaluated the performance of our model in terms of accuracy with 246 test data. Furthermore, three forensic scientists evaluated the 3D skull images, and their performances were compared with those of the model. Our model showed an accuracy of 0.93, which was higher than that of the forensic scientists. Our model primarily focused on the entire skull owing to visualization but focused less on the areas often investigated by forensic scientists. In summary, our model may serve as a supportive tool to identify cadaver sex based on skull shape. Further studies are required to improve the model's performance.

Deep Neural Networks-Based Age Estimation of Cadavers Using CT Imaging of Vertebrae.

Although age estimation upon death is important in the identification of unknown cadavers for forensic scientists, to the best of our knowledge, no study has examined the utility of deep neural network (DNN) models for age estimation among cadavers. We performed a postmortem computed tomography (CT) examination of 1000 and 500 male and female cadavers, respectively. These CT slices were converted into 3-dimensional images, and only the thoracolumbar region was extracted. Eighty percent of them were categorized as training datasets and the others as test datasets for both sexes. We fine-tuned the ResNet152 models using the training datasets. We conducted 4-fold cross-validation, and the mean absolute error (MAE) of the test datasets was calculated using the ensemble learning of four ResNet152 models. Consequently, the MAE of the male and female models was 7.25 and 7.16, respectively. Our study shows that DNN models can be useful tools in the field of forensic medicine.