Southeastern Health AI Consortium Grant Awardees
Advancing AI-driven discovery workflows is a core part of the University of Florida Clinical and Translational Science Institute’s (UF CTSI) strategy, and its newest cross-institutional effort puts that strategy into action. Through the Southeastern Health AI Consortium (SEHAI), a partnership with the Medical University of South Carolina (MUSC), UF CTSI has helped launch $6.6 million in pilot awards designed to test how AI can move from a promising tool to a practical driver of better health outcomes across the Southeast.
UF CTSI Co-Director Dr. Elizabeth A. Shenkman is helping lead the initiative, and her vision reflects the institute’s broader goal: proving AI’s value in real clinical workflows, not just in theory. “Maternal morbidity and mortality due to hypertensive disorders of pregnancy can potentially be reduced through more rapid identification of high-risk mothers,” Shenkman said. “The partnership and associated data assets provide a unique opportunity to develop algorithms to identify those at risk so that interventions can be implemented earlier.”
The awards are structured to generate exactly that kind of evidence, through two tracks. Novel Health Insights is funding six teams with up to $400,000 over two years to close real-world care gaps and build the preliminary data needed to compete for larger federal funding. Methodologic Innovation is backing two teams with up to $100,000 in one-year seed funding to pilot new methods in health services research, biomedical informatics and data science. Each project pairs a UF investigator with a MUSC counterpart, and each incorporates AI directly or lays the groundwork for it.
The eight funded projects cluster around the challenges SEHAI was built to tackle. On maternal health, UF CTSI’s own Dr. Yonghui Wu, chief data scientist, is applying large language models to improve maternal mental health screening, while Dr. Dominick Lemas is developing AI-driven “digital twins” to model pregnancy interventions before they happen. On the mental and behavioral health side, Dr. Joni Splett is using AI to strengthen youth mental health referrals in schools, and Dr. Qianqian Song is applying AI to improve depression care for patients facing likely incurable cancer. A third cluster is focused on brain health and neurodegenerative disease: Dr. Glenn Smith is building outcome-prediction models that adapt across diverse patient groups, Dr. Nikolaus McFarland is developing automated speech and language assessments for earlier diagnosis of neurodegenerative disorders, and Dr. Jie Xu is building privacy-preserving, federated-learning approaches to advance gene and cellular therapies and brain health research. Rounding out the portfolio, Dr. Gilbert Upchurch Jr. is pairing imaging and demographic data with AI to sharpen stroke risk prediction and personalize carotid disease management.
For UF CTSI, SEHAI is a direct test of what the institute’s AI-impact discovery pipeline can look like at scale: pairing UF’s data and researchers with an outside institution to see which AI applications actually change care and which need more work before they’re ready for it. “The MUSC-UF partnership is uniquely positioned to lead AI-enabled research because of the breadth of our patient populations and the range of communities within which care is delivered,” Shenkman said — a fitting summary of why this pilot matters, and where UF CTSI hopes to take it next. Read the full article from MUSC here.
“Improving Maternal Mental Health Screening Using Large Language Model-based Artificial Intelligence”
PI: Yonghui Wu, Ph.D.
Professor and Chief Data Scientist
Wu directs Natural Language Processing at UF CTSI, and his research centers on large language models, machine learning, and computational drug discovery using electronic health records. He’s built a number of widely used AI models, including the GatorTron large language models, and has served as principal investigator on grants from the NIH and the Patient-Centered Outcomes Research Institute. He earned his Ph.D. in Computer Science from the Harbin Institute of Technology and trained in biomedical informatics at Vanderbilt University and the University of Texas Health Science Center at Houston. On this project, he’s partnering with Constance Guille, M.D., professor of Psychiatry and Behavioral Sciences at the Medical University of South Carolina (MUSC).
“Transforming Depression Treatment for Individuals with Likely Incurable Cancer: Electronic Health Record Large Language Model-Based Identification”
PI: Qianqian Song, Ph.D.
Assistant Professor and Director of Bioinformatics
Song is using large language models to identify depression directly from the electronic health records of patients facing likely incurable cancer — a step toward catching and treating depression in this population earlier and more reliably. She directs the Cancer Data Commons and Translational Bioinformatics at the UF Health Cancer Center, where she also co-leads the Cancer AI Working Group and the Computational Biology Unit. Her research centers on building AI and machine learning tools that integrate large-scale biomedical data — from genomic and single-cell data to electronic health records — to better understand disease and identify new treatment targets. On this project, she’s partnering with Evan Graboyes, M.D.., professor of Otolaryngology at MUSC.
“Automated Speech and Language Assessment to Improve Diagnosis of Neurodegenerative Disorders”
PI: Nikolaus McFarland, M.D., Ph.D.
Clinical Professor of Neurology
McFarland is a movement disorders specialist at the Norman Fixel Institute for Neurological Diseases and holds the Wright/Falls/Simmons Professorship in Progressive Supranuclear Palsy (PSP)/Atypical Parkinson’s. He directs a multidisciplinary clinical research program for atypical Parkinson’s disorders — recognized as a CurePSP Center of Care — as well as the UF Huntington Disease Society of America Center of Excellence. His research focuses on understanding the mechanisms behind Parkinson’s disease and related disorders, and on identifying novel biomarkers for earlier detection, tracking, and treatment. On this project, he’s partnering with Federico Rodriguez-Porcel, M.D.., associate professor of Neurology at MUSC.
“Precision Pregnancy Digital Twins: A Collaborative Pilot for AI-Driven Intervention Modeling”
PI: Dominick Lemas, Ph.D.
Assistant Professor of Health Outcomes and Biomedical Informatics
Lemas’s research is devoted to understanding the fetal origins of pediatric obesity, with a particular focus on gut microflora and the host-microbe interactions that shape maternal-infant metabolism. His broader interests span biomedical informatics, computational biology, and molecular epidemiology as they relate to maternal-child health. He earned his Ph.D. in biochemistry and molecular biology from the University of Alaska Fairbanks. On this project, he’s partnering with Kelly Hunt, Ph.D., professor of Public Health Sciences at MUSC.
“Integration of Demographics and Medical Imaging Utilizing Artificial Intelligence and Fluid Dynamics to Improve Stroke Risk Prediction and Develop Personalized Carotid Disease Management”
PI: Gilbert Upchurch Jr., M.D., DHA
Professor of Vascular Surgery
Upchurch is the Edward M. Copeland III and Ann & Ira Horowitz Chair of the UF Department of Surgery, and an internationally recognized clinician, researcher, and educator in the treatment of aortic and vascular disease. His current research focuses on better understanding the pathogenesis of aortic aneurysms, and his clinical expertise centers on treating thoracoabdominal aortic aneurysms. He was inducted into the National Academy of Medicine in 2021. On this project, he’s partnering with Ravi Veeraswamy, M.D.., professor of Surgery at MUSC.
“A Novel Approach for Building and Optimizing Outcome Prediction Models for Multiple-Heterogeneous Cohorts: A Case Study Using a Multicomponent Intervention in Mild Cognitive Impairment”
PI: Glenn Smith, Ph.D.
Professor of Neuropsychology
Smith serves as contact PI of the 1Florida Alzheimer’s Disease Research Center and is a board-certified neuropsychologist who has practiced clinically for more than 30 years. His research focuses on the early detection of neurodegenerative disease, with more than 25 years spent studying mild cognitive impairment, alongside behavioral interventions designed to prevent or delay dementia. He developed and leads the HABIT® (Healthy Action to Benefit Independence and Thinking) program, an intensive behavioral intervention for people with mild cognitive impairment and their care partners. On this project, he’s partnering with Stephanie Aghamoosa, Ph.D., assistant professor of Neuropsychology at MUSC.
“AI-Enhanced Team Monitoring to Improve Youth Mental Health Referrals in Schools”
PI: Joni Splett, Ph.D.
Associate Professor of Psychology
Splett directs the University of Florida Prevention and Intervention Network and is a licensed psychologist whose work focuses on systems-level strategies for youth mental health promotion, prevention, and intervention in schools — including universal mental health screening and team-based approaches like the Interconnected Systems Framework. Her research has been supported by the National Institutes of Health, the Institute of Education Sciences, and the U.S. Department of Education, among others. On this project, she’s partnering with Colleen Halliday, Ph.D., associate professor of Psychiatry and Behavioral Sciences at MUSC.
“A Portfolio of Use Cases to Enable Privacy-Preserving AI with Federated Learning: Applications to Advance Gene and Cellular Therapies and Address Brain Health”
PI: Jie Xu, Ph.D.
Assistant Professor of Health Outcomes and Biomedical Informatics
Xu’s research sits at the intersection of machine learning and health informatics, with particular expertise in federated learning techniques that allow algorithms to be trained across multiple institutions without sharing underlying patient data — directly relevant to this project’s privacy-preserving approach. She also works on disease progression subtyping and predictive risk modeling using electronic health records, claims data, and medical imaging. Before joining UF, she was a postdoctoral researcher at Weill Cornell Medicine, and she earned her Ph.D. in Electrical Engineering from Xidian University. On this project, she’s partnering with Paul Heider, Ph.D., assistant professor of Public Health Sciences at MUSC.