Tohoku J. Exp. Med., 2023 July, 260(3)
Deep Learning-Based Diagnosis of Fatal Hypothermia Using Post-Mortem Computed Tomography
Yuwen Zeng,1 Xiaoyong Zhang,2,3 Issei Yoshizumi,4 Zhang Zhang,1 Taihei Mizuno,5 Shota Sakamoto,5 Yusuke Kawasumi,4 Akihito Usui,4 Kei Ichiji,4 Ivo Bukovsky,6,7 Masato Funayama4 and Noriyasu Homma4
1Department of Intelligent Biomedical Systems Engineering, Graduate School of Biomedical Engineering, Tohoku University, Sendai, Miyagi, Japan
2National Institute of Technology, Sendai College, Sendai, Miyagi, Japan
3Institute of Development, Aging and Cancer, Tohoku University, Sendai, Miyagi, Japan
4Department of Radiological Imaging and Informatics, Tohoku University Graduate School of Medicine, Sendai, Miyagi, Japan
5Department of Management Science and Technology, Graduate School of Engineering, Tohoku University, Sendai, Miyagi, Japan
6Faculty of Science, University of South Bohemia in Ceske Budejovice, Ceske Budejovice, Czech Republic
7Faculty of Mechanical Engineering, Czech Technical University in Prague, Prague, Czech Republic
AIn forensic medicine, fatal hypothermia diagnosis is not always easy because findings are not specific, especially if traumatized. Post-mortem computed tomography (PMCT) is a useful adjunct to the cause-of-death diagnosis and some qualitative image character analysis, such as diffuse hyperaeration with decreased vascularity or pulmonary emphysema, have also been utilized for fatal hypothermia. However, it is challenging for inexperienced forensic pathologists to recognize the subtle differences of fatal hypothermia in PMCT images. In this study, we developed a deep learning-based diagnosis system for fatal hypothermia and explored the possibility of being an alternative diagnostic for forensic pathologists. An in-house dataset of forensic autopsy proven samples was used for the development and performance evaluation of the deep learning system. We used the area under the receiver operating characteristic curve (AUC) of the system for evaluation, and a human-expert comparable AUC value of 0.905, sensitivity of 0.948, and specificity of 0.741 were achieved. The experimental results clearly demonstrated the usefulness and feasibility of the deep learning system for fatal hypothermia diagnosis.
Key words —— artificial intelligence; autopsy; deep learning; fatal hypothermia; post-mortem computed tomography
© 2023 Tohoku University Medical Press
===============================
Tohoku J. Exp. Med., 2023 July, 260(3), 253-261.
Correspondence: Yuwen Zeng, Department of Intelligent Biomedical System Engineering, Graduate School of Biomedical Engineering, Tohoku University, 2-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-0872, Japan.
e-mail: zeng.yuwen.s4@dc.tohoku.ac.jp