Owsley and McGwin will lead three data collection sites in collecting data to inform machine learning approaches to provide critical insights into the endemic condition of type 2 diabetes mellitus.
research team members from a wide variety of backgrounds and disciplines to generate ethically-sourced tools, data and resources, and ultimately bridge the gap between biomedical and behavioral research and artificial intelligence. .The NIH Common Fund’s Bridge to Artificial Intelligence program, or Bridge2AI, will help set the stage for the widespread use of artificial intelligence to solve some of the most pressing challenges in human health. The program will bring together
Bridge2AI is a trans-NIH program funded by the NIH Common Fund that is collaboratively overseen by the NIH Common Fund, National Center for Complementary and Integrative Health, National Eye Institute, National Human Genome Research Institute, National Institute of Biomedical Imaging and Bioengineering and the National Library of Medicine.
Two research teams led by the University of Alabama at Birmingham out of approximately 100 teams competing nationally were selected for funding through this program. One will be led by the Department of Ophthalmology and Visual Sciences and the other by researchers from the UAB Institute of Computer Science.
The UAB Department of Ophthalmology and Visual Sciences team received a one-year, $2 million grant from the NIH’s Bridge2AI program (OT2OD032644). The program is funded on an annual basis for each of the four years. The grant is based at the University of Washington, whose total first-year award is $7.8 million.
The UAB stands for a module called Data Acquisition, the most important of the six modules awarded by the University of Washington.
Cynthia Owsley, Ph.D., professor and director of the Clinical Research Unit in the Department of Ophthalmology and Visual Sciences at UAB Marnix E. Heersink School of Medicine, and Gerald McGwin Jr., Ph.D ., professor in the UAB School of Public Health, in collaboration with Jeffrey Edberg, Ph.D., professor in the UAB Department of Medicine, will oversee the creation of flagship datasets. These datasets will be based on ethical principles, standards and associated tools, as well as skills and workforce development. The goal is to address major challenges in biomedical and behavioral research that require artificial analysis and machine learning.
“Our award is to lead three sites in collecting data that will be used for machine learning approaches that will provide critical insights into the endemic condition of type 2 diabetes mellitus,” Owsley said.
“Our approach is to collect data on 4,000 adults in three regions of the United States with equal representation of four racial/ethnic groups and stages of type 2 diabetes severity,” Owsley said. “Creating balanced training datasets is essential for the development of unbiased AI/ML models. So rather than targeting the demographic distribution of the US population, we will intentionally recruit an equal number of four racial groups /ethnic – Blacks, Whites, and those of Asian descent and those of Hispanic descent.
Owsley adds that the same reasoning applies to balancing the severity of diabetes, pre-diabetes, lifestyle-controlled, oral medication-controlled, and insulin-controlled diabetes. AI/ML-ready data will include social determinants of health, continuous glucose monitoring, serological testing for endocrine, cardiac and kidney biomarkers, genome-wide polymorphism assessment, imaging visual and retinal, cognitive tests, electrocardiogram and 24-hour activity monitoring.
Data collection sites will be at UAB, University of California San Diego, and University of Washington. The UAB Center for Clinical and Translational Sciences Biological Specimen Repository will oversee biological specimen specimens for all sites.