Batool, Syeda Naila and Yang, Jing and Gilanie, Ghulam and Latif, Akkasha and Yasin, Shumaila and Ikram, Amna and Por, Lip Yee (2025) Forensic Radiology: A robust approach to biological profile estimation from bone image analysis using deep learning. Biomedical Signal Processing and Control, 105. p. 107661. ISSN 1746-8094, DOI https://doi.org/10.1016/j.bspc.2025.107661.
Full text not available from this repository.Abstract
Human identification is of paramount importance in today's world, especially in clinical contexts. Identifying a person solely from their bones presents an exceptionally challenging task. Natural disasters and incidents can lead to the loss of identity. In such cases, courts, police, and other law enforcement organizations turn to forensic departments to estimate crucial details like age, gender, and height. However, this process is not only timeconsuming but also costly. In this research work, employ computer vision-based techniques to identify age and gender from bone X-ray images. This work utilizes the morphological and deep texture characteristics of these images for biological profile estimation. This proposed approach achieved an accuracy of 75.34 % on one test set and an accuracy of 94.98 % on another test set, as measured by standard evaluation criteria, which encompassed 3206 images from diverse age and gender groups. To the best of our knowledge, this study is the first of its kind in Pakistan. It has the potential to assist the police, criminologists, and other law enforcement agencies in conducting quicker, more feasible, expedited, and cost-effective examinations. This study also compares results with the gold standard method, DNA analysis, for bone investigations.
| Item Type: | Article |
|---|---|
| Funders: | UNSPECIFIED |
| Uncontrolled Keywords: | Forensic Radiology; Biological Profile Estimation; Skeleton-based Forensic Radiology; Convolutional Neural Network; Bones X-ray |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Faculty of Computer Science & Information Technology > Department of Computer System & Technology |
| Depositing User: | Ms. Juhaida Abd Rahim |
| Date Deposited: | 29 Apr 2025 04:49 |
| Last Modified: | 29 Apr 2025 04:49 |
| URI: | http://eprints.um.edu.my/id/eprint/47911 |
Actions (login required)
 |
View Item |
