A systematic review about the evolving role of artificial intelligence in various fields of forensic medicine

This study systematically reviews the applications and impact of artificial intelligence (AI) in forensic medicine, focusing on its role in mimicking human cognitive processes, enhancing diagnostic accuracy, pattern recognition, and operational efficiency across forensic domains.

A systematic search was conducted in PubMed and Google Scholar using keywords including "AI in forensics" and "machine learning forensic analysis," covering publications from 2014 to 2024. A total of approximately 1000 articles were initially identified, of which 100 met the inclusion criteria after screening for relevance, study design, and quality.

AI applications in forensic medicine were categorized into key domains: personal identification, forensic pathology, radiology and imaging, digital forensics, toxicology, and forensic anthropology. Machine learning, deep learning, and neural network models demonstrated improvements in accuracy, reproducibility, and efficiency compared with conventional approaches. For example, AI-assisted imaging techniques reduced inter-observer variability in postmortem fracture detection, while predictive models for postmortem interval estimation showed mean error reductions of up to 15 %. Despite these advances, challenges such as small, non-representative datasets, limited external validation, and ethical concerns remain.

AI has significantly enhanced multiple areas of forensic practice by improving diagnostic capabilities, streamlining workflows, and supporting decision-making. However, wider adoption requires rigorous validation, standardization, and ethical oversight. Future research should focus on integrating multimodal data, expanding dataset diversity, and addressing legal and ethical implications to maximize the utility of AI in real-world forensic investigations.