![]()
Available research evidence suggests that artificial intelligence (AI) approaches could play clinically useful roles in occupational medicine, including health risk assessment and prediction of disability duration and return to work, according to a review article in the December
Journal of Occupational and Environmental Medicine.
Zaira S. Chaudhry, MD, MPH, and Avishek Choudhury, PhD, of West Virginia University, Morgantown, performed a systematic review of studies on clinically relevant applications of AI in occupational health, published between 2014 and 2024. Qualitative analysis focused on 27 studies using a total of 47 AI algorithms, including artificial neural networks, support vector machines, and random forest.
Seventeen studies focused on occupational health risk assessment. Findings included positive results using AI to predict noise-induced hearing loss in steel factory workers and changes in white blood cell counts among workers exposed to benzene. Other papers reported positive evaluations of AI in risk assessment for musculoskeletal disorders, occupational lung disease, blood dyscrasias, and metabolic syndrome.
Five studies examined the use of AI in predicting return to work and disability duration—uniquely important factors in assessment and management of workers' compensation cases. Other studies evaluated clinical applications of AI in assessment of injury severity and development of occupational injury management decision support tools and mobile health applications.
"While the findings of the reviewed studies are promising, it is necessary to proceed with caution when integrating AI into occupational health settings," Drs. Chaudhry and Choudhury conclude. They identify key areas for further research—highlighting the need for "robust explainable AI models that are informed by occupational health clinicians and rigorously validated in real-world settings with diverse worker populations to optimize their clinical utility and promote clinician trust in these models."