by John W. Mitchell
, Senior Correspondent | September 11, 2018
Researchers at The American College of Radiology’s Neiman Institute and the Georgia Institute of Technology will apply big data analytics and artificial intelligence (AI) not to images, but the underlying imaging claims data.
The goal of the $3 million, five-year project is to unveil new insights on how healthcare imaging delivery and payments affect providers and patients. The project will be dubbed HEAL, for Health Economics and Analytics Lab, and will be housed within Georgia Tech’s Ivan Allen College of Liberal Arts.
“In the same way that big data and AI can transform the process of interpreting medical images, applying these tools to large medical claims databases can reveal hidden relationships and nonintuitive implications of different health policies, which can be used to better design care for patients, providers, and payers,” Dr. Danny Hughes (Ph.D.), executive director and senior research fellow told HCB News. “To our knowledge, we are the only group leveraging these tools in a policy-oriented way.”
The mission of the Neiman Institute is not centered on the clinical aspect of imaging, but the specialty’s role in evolving health care delivery and payment models, according to Hughes. As an example of the kind of insight he hopes the HEAL partnership will achieve, he cited a mammography question.
“[Could] changing mammography screening from a fee-for-service to bundled payment reimbursement [have] the potential to substantially increase adherence to screening guidelines, which potentially has profound, positive implications for population health?” Hughes asked.
The Nieman Institute will provide HEAL with its extensive data resource of medical claims covering millions of American lives. HEAL will support full-time post-doctorate researchers, graduate research assistants, and affiliated Georgia Tech faculty to produce methodical and policy-oriented research.
“Aside from the clinical perspective, access to large volumes of medical claims data, such as those now available, can unlock relationships that weren't observable with traditional analyses on the relatively small datasets of the past,” Hughes said.
The ability to observe millions of care pathways and resulting outcomes should provide tremendous insights into the best methods to manage care, according to Hughes. He added that HEAL is a step closer toward the promise of personalized health care to optimize outcomes for each patient.