Initiative seeks to improve diversity of AI researchOct. 21, 2021, 10:09 AM
by Paul Govern
This summer the National Institutes of Health (NIH) announced a new flagship initiative, the Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity, or AIM-AHEAD.
The initiative seeks to quell common biases that beset biomedicine and aims specifically to redress two glaring deficits identified by the NIH: a lack of diversity among artificial intelligence researchers and a lack of representation from various segments of society in the data available to support machine learning and the implementation of artificial intelligence in the health domain.
To get the initiative going, on Sept. 22 the NIH announced a $50 million award to the University of North Texas Health Science Center (UNTHSC) to establish the AIM-AHEAD Coordinating Center. The NIH will spend an additional $50 million on the initiative next year.
The Coordinating Center includes 14 investigators from around the country, among them Vanderbilt University Medical Center’s Toufeeq Ahmed, PhD, MS, assistant professor of Biomedical Informatics, who’s one of nine investigators for the center’s Leadership Core, and Bradley Malin, PhD, Accenture Professor of Biomedical Informatics, Biostatistics and Computer Science, who’s one of three investigators for the center’s Infrastructure Core. The lead investigator for the Coordinating Center is Jamboor Vishwanatha, PhD, of UNTHSC.
“Within the Leadership Core, along with administering the project as a whole, we’ll focus on recruiting and building a consortium of partners to enhance both the inclusivity of health data used for machine learning and the diversity of leadership in the field of artificial intelligence as it applies to biomedicine,” Ahmed said.
Anil Shanker, PhD, from Nashville’s Meharry Medical College, is also among investigators in the Leadership Core. With UNTHSC figuring as the central hub, the Leadership Core will establish five regional hubs for the consortium.
Ahmed’s team at VUMC will be responsible for web-based communications capabilities used to build the consortium and advance its agenda. An AIM-AHEAD website, designed by Ahmed, was launched recently.
Working with Malin on the Infrastructure Core are the core’s lead investigator, Alex Carlisle, PhD, from the National Alliance Against Disparities in Patient Health in Woodbridge, Virginia, and Paul Avillach, MD, PhD, from Harvard University in Boston.
“The Infrastructure Core will first conduct a capabilities assessment to determine which types of platforms and systems will best support the sustainable application of artificial intelligence and machine learning methods to the study of health disparities and the training of investigators to continue this work in the future,” Malin said. He noted that VUMC has experience in building infrastructure to support a wide range of systems, including the All of Us Research Program, the STAR Clinical Data Research Network, the Electronic Medical Records and Genomics Network and the National Research Mentoring Network.
Data used by AIM-AHEAD are anticipated to range from electronic health records to medical images, genomic data, socio-economic data drawn from public sources, and environmental factors.
“Once our assessment is complete, we’ll focus on getting these data curated and ready for research. In full coordination with the other cores at AIM-AHEAD, we’re out to build a scalable, flexible and sustainable research environment that will provide efficient, cost-effective access to qualified researchers,” Malin said.
Rounding out the AIM-AHEAD Coordinating Center are two additional components — the Data Science Training Core, led by an investigator at Howard University in Washington, D.C., and the Data and Research Core, led by an investigator at OCHIN, formerly known as the Oregon Community Health Information Network.
The AIM-AHEAD initiative is funded and directed by the NIH (OT2OD032581).