UMass Chan to open AI assurance lab to validate use of the technology in health care
Image I Courtesy of UMass Chan Medical School
Industry leaders gathered at UMass Chan Medical School in Worcester on Wednesday for the announcement of the Health AI Assurance Laboratory were (from left) Francis Campion, MITRE; Ozgur Eris, MITRE; Brian Anderson, the Coalition for Health AI; Carolyn Kirk, Massachusetts Technology Collaborative; Patrick Larkin, MassTech; Mass. Economic Development Secretary Yvonne Hao; Terence Flotte, UMass Chan; Michael Collins, UMass Chan; state Sen. Robyn Kennedy (D-Worcester); Doug Robbins, MITRE; Melissa Fischer, UMass Chan; Miles Thompson, MITRE; and David McManus, UMass Chan.
A great deal of concern surrounds the applications of artificial intelligence in health care, said Dr. David McManus, chair and professor of medicine at UMass Chan Medical School in Worcester. Yet, he said, the technology is already here, and that’s where his new lab comes into play.
On Wednesday, UMass Chan announced the creation of the university’s Health AI Assurance Laboratory: a lab designed to evaluate processes and impacts of health AI technology to advance efforts for more accountable, unbiased, and efficacious patient care. The lab is being established in partnership with Bedford non-for-profit research-and-development center MITRE.
“At present, within our own UMass health ecosystem, we have myriad different AI tools currently in use, and that's true across almost every regional community and academic medical center. But there are relatively few labs available to actually evaluate the both validity but also the fairness, the usability, and some of the workflow challenges that come from trying to apply these [AI] tools,” McManus said at the Wednesday press conference.
The lab will be the first of its kind in Massachusetts, utilizing a human-in-the-loop testing model to address concerns including technical robustness and safety, transparency, diversity and fairness, and society and environmental well-being when it comes to AI technology, according to a Wednesday press release from UMass Chan. Evaluation guidelines will be established by MITRE’s AI Assurance and Discovery Lab in Virginia and the Coalition for Health AI, located in Boston.
The lab will be composed of three main components: the physical space used to test the application of health AI; the design and implementation of the health AI knowledge base, which is supplemental to the Health AI Assurance Lab; and the development of the next generation of a health AI workforce.
Leading the UMass lab with McManus are Dr. Adrian Zai, UMass Chan associate professor; Miles Thompson, AI assurance solutions lead at MITRE,. This team, along with professionals across education, health care, private industry, nonprofit, and public sectors, will work together to assess specific health AI applications.
“Our mission is to grow that regional trustworthy health AI industry by empowering this ecosystem, and we want to bring in tools with industry partners and also train the next generation of both healthcare workers and also engineers to develop tools with fairness built in from the start,” McManus said at the event.
The lab has received a $555,000 Technology & Innovation Ecosystem Awards Program grant from the Innovation Institute at the Massachusetts Technology Collaborative.
“We’re excited to continue to invest, to lengthen our lead, and to solve big problems that are going to help everybody in our state, hopefully in the country, and in the world,” Yvonne Hao, Massachusetts secretary of economic development, said in a WBJ interview following the Wednesday announcement. “We know that this investment is going to pay off in spades, and we’re going to actually advance this work to help think about how we can use AI in positive, good ways to solve health care problems.”
The lab is receiving $137,000 in matching private investments.
Dr. Brian Anderson, CEO of CHAI, likens the lab to an organization like New York nonprofit Consumer Reports.
“To be able to validate these models and partner with the innovators to potentially remediate a model where they might be bias,” said Anderson. “This is the vision for that. These are the assurance labs that are going to be the future.”
WBJ Web Partners
Most Recent
Most Recent
0 Comments