BROOKINGS — South Dakota State University has been selected to lead a research project on the National Institutes of Health-funded Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity (AIM-AHEAD) program coordinated by the University of North Texas Health Science Center at Fort Worth. The project’s title is “Developing Explainable Machine Learning and Computational Methods for Identifying Geographic and Racial Disparities in End-Stage Renal Disease.”

The research team is led by principal investigator Semhar Michael, associate professor of statistics in SDSU’s Jerome J. Lohr College of Engineering. The project will analyze the racial and geographic disparities in mortality with end-stage renal (kidney) disease with a focus on Native American and Hispanic populations in South Dakota.

“This project will develop computational and machine learning models to assess disparities in end-stage renal disease,” Michael said. “This is a collaborative effort that involves interdisciplinary team members from statistics, nursing, pharmacy, health informatics and health care providers.”

This two-year, $1,091,316 project will also identify factors associated with the risk of mortality for these two population groups. The team, which includes Michael and SDSU assistant professors Hossein Moradi Rekabdarkolaee (Lohr College of Engineering) and Brandon Varilek (College of Nursing), will utilize machine learning and computational models to develop the predictive tool. The SDSU team will collaborate with Patti Brooks, a faculty member from Dakota State University, and Surachat Ngorsuraches, a faculty member from Auburn University, as well as healthcare partners from Avera Medical Group Nephrology, Great Plains Tribal Leaders’ Health Board and Sanford Health.

In the United States, Native Americans and Hispanics are more likely to die from end-stage renal disease than other population groups, according to the United States Renal Data System. Native Americans and Hispanics also experience increased health disparities, primarily due to their social determinants of health. This project will develop a model that will predict the risk of mortality among Native Americans and Hispanics with end-stage renal disease who have varying socio-demographic characteristics and social determinants of health.

“We hope to deliver a fair and unbiased tool that can predict the risk of mortality for a person with end-stage renal disease given their socio-demographic, economic status, location, access to healthcare facilities, etc.” Michael said. “The goal ultimately is to update the developed framework to an individual level to provide the opportunity for healthcare providers and policymakers to mitigate the health disparities among Native American and Hispanic persons with ESRD in South Dakota and possibly other regions.”

The project is supported by the National Institutes of Health (NIH) Agreement No. 1OT2OD032581.