At many American hospitals, bedside monitors that measure everything from patients’ intracranial pressure to heart, respiration and blood pressure rates feed numbers into artificial intelligence systems that constantly assess their risk of suffering things like stroke, sepsis and heart attack.
But because the algorithms that feed such predictive health care AI systems are often based on data from homogenous populations, AI that’s meant to improve care for everyone sometimes falls short – and may even prove harmful.
Supported by a new $5.9 million National Institutes of Health grant, two University of Virginia researchers will explore ways to improve the use of artificial intelligence in health care for a wider diversity of patient populations.
Primary co-investigators Ishan Williams, an associate professor in the School of Nursing, and Randall Moorman, a UVA Health cardiologist, will develop, test and deploy best practices for artificial intelligence health care systems that aggregate data from a more diverse pool of patients by taking into account their race, ethnicity, socio-economic status and geography.
In a nation with growing racial and ethnic diversity, the effort is a priority for the NIH, which, with its “Bridge to AI” programs, has assembled an array of scholars tasked with improving artificial intelligence.
COVID-19 has proven particularly vexing for communities of color, which have been devastated by the disease, compared to their white peers. “If COVID has taught us anything,” Williams said, “it’s that we must take a nuanced, inclusive, equitable approach to everything we do in health care, AI included.”
Acting as co-primary investigators and module leaders, Moorman and Williams will work with investigators at Massachusetts General Hospital, the University of Florida, and colleagues across the country in biomedical and clinical organizations, industry and regulatory agencies.
The research could ultimately help increase AI’s ability to play a more effective role in the care of diverse patients.
Williams’s and Moorman’s work began in late 2022; they expect the AI collaborative, training materials, and diverse data sets to be completed by 2027.