AI models are being used to track zoonotic diseases. Will they prevent the next pandemic?

Bwindi Impenetrable National Park, located in southwestern Uganda, is renowned for its endangered mountain gorillas and diverse wildlife. However, its proximity to human settlements and livestock increases the risk of spillover events, where pathogens can transfer between species. In 2003, Conservation Through Public Health (CTPH) began a holistic One Health approach to protect gorillas from diseases originating in humans.

Recently, CTPH has shifted towards a more proactive approach to disease prevention, utilizing artificial intelligence (AI) to train predictive AI models that could serve as early-warning systems for potential outbreaks affecting people, livestock, and gorillas.

Zoonotic diseases, which originate in animals, pose a significant threat to human health, as seen in cases like SARS-CoV-2 and ebolaviruses. Factors such as agriculture, climate shifts, deforestation, and globalization have contributed to the increased likelihood of spillover events. AI tools have the potential to help manage zoonotic diseases and reduce the risk of spillover, but they are still limited and not widely used in infectious-disease epidemiology.

Ronald Ogwal, a public-health specialist at CTPH, has been collecting samples from cattle and poultry around Bwindi to test for diseases like brucellosis and Rift Valley fever. The results are shared with local communities and integrated with CTPH's routine gorilla-monitoring records. These data are then used in AI models designed to identify and predict potential disease outbreaks, serving as an early-warning system.

AI tools like machine-learning models and deep-learning algorithms are being developed to recognize traits associated with disease emergence and to identify "risky" viruses that researchers can further investigate for vaccination purposes.

While AI models are primarily being used for research purposes at the moment, there is potential for them to be integrated into surveillance efforts. Regular sampling for specific antibodies could be conducted in key areas where spillover is likely to occur, such as live animal markets, poultry farms, or settlements near bat roosts. This proactive approach could help mitigate the risks of zoonotic disease outbreaks and potentially prevent future pandemics.