A Scoping Review of Machine Learning Applications Across Epidemiological Stages of Zoonotic Disease

Emerging zoonotic diseases represent a significant threat to global health. While machine learning (ML) holds promise for their management, a comprehensive understanding of how these technologies are applied across the entire animal-to-human transmission pathway is lacking. This scoping review systematically maps ML applications in zoonotic disease management to identify research trends, methodological approaches, and critical gaps across different epidemiological stages and functional domains. We organize the literature along two dimensions: epidemiological stages, from animal hosts to human populations, and functional domain, including diagnosis, epidemiology, and intervention. We searched PubMed and Web of Science for studies on 14 preselected high-priority zoonotic diseases. The search string combined keywords for the selected diseases, ML techniques, and functional applications (diagnosis, epidemiology, and intervention). A total of 966 studies were included in the final analysis, of which 72.8% focused on COVID-19. Our analysis shows robust ML performance in clinical diagnostics, epidemic forecasting, and intervention optimization within human populations. However, critical gaps persist, only 1.96% of studies examined the animal-human interface, no ML models explicitly targeted spillover prevention, and studies on animal-reservoir surveillance remain limited. All spillover studies originated from high-income or upper-middle-income countries (UMICs), in contrast with low- and lower-middle-income countries (LMICs) contributing 21.4% of human-stage studies. These findings reveal a pronounced mismatch between research investment and spillover risk and highlight the need for greater emphasis on spillover mechanisms, enhanced integration of cross-species transmission dynamics, and methods suitable for surveillance in resource-limited settings. Addressing these imbalances is essential for advancing a shift from reactive outbreak response to proactive spillover prevention within a One Health framework.