AI Technology Enhances Rapid Identification in Forensic Cases Using Chest Radiographs

A collaborative effort from Michigan State University's Department of Anthropology and Computer Science and Engineering has led to the development of an innovative AI-powered method to assist forensic anthropologists in human identification. By analyzing over 5,000 chest radiographs, the study employed deep neural networks—advanced AI algorithms—to identify specific regions of interest within the images that are critical for identifying individuals. This breakthrough enables forensic teams to quickly shortlist potential matches, significantly reducing time-consuming manual comparisons.

In mass fatality scenarios, where rapid identification of numerous individuals is essential, this AI system can analyze more than 1,800 radiographs in just 17 seconds—a task that would otherwise take human practitioners 30 to 60 hours. Such efficiency improvement not only expedites case processing but also helps mitigate practitioner bias when matching unidentified remains against missing persons databases.

Moreover, this technology marks the first application to evaluate how different regions within chest radiographs can be utilized for personal identification in forensic contexts, highlighting a distinct approach compared to traditional medical uses which focus on disease diagnosis. The interdisciplinary team, including researchers like Dr. Carolyn Isaac and Dr. Arun Ross, emphasizes the unique benefits of combining computer science expertise with forensic anthropology, fostering innovative solutions for difficult forensic challenges.

Published in IEEE Access, this study underscores the potential of AI to transform forensic investigations by providing faster, more reliable identification methods, and paves the way for future developments in digital forensic tools.