fb-pixel Skip to main content

New research center at UMass Amherst will use Artificial Intelligence to improve at-home care for elderly patients

W. E. B. Du Bois Library and the Integrative Learning Center can be seen across the Campus Pond at UMass Amherst.Erin Clark/Globe Staff

What’s next for the ever-developing industry of heartbeat-monitoring wristwatches and voice-responsive phones? According to the leaders of a newly endowed Massachusetts research center, the devices’ built-in Artificial Intelligence could prove useful in improving the quality of at-home care for the elderly.

Unveiled this week, the Massachusetts AI and Technology Center for Connected Care in Aging and Alzheimer’s Disease, based at the University of Massachusetts Amherst, will work with a confluence of new and existing technologies, drawing on AI to modernize the at-home care industry for those with age-related ailments and Alzheimer’s Disease patients.

The project seeks to address what its founders see as “a major healthcare disparity” that can leave the elderly with a vexing choice: stay home and receive a lower level of care, or leave home for proper treatment.


“More than 90% of older Americans would prefer to stay in their homes as they age,” a press release announcing the new center said. “However, the prevalence of chronic illness, including Alzheimer’s disease, can make the goal of successful aging at home out of reach without substantial support.”

Computer scientists and doctors from Brigham and Women’s Hospital, Massachusetts General Hospital, Brandeis University, and Northeastern University will partner on the research, which will be funded by roughly $20 million in grants from the National Institute on Aging distributed over the next five years.

“Artificial intelligence has the potential to transform important areas of science and medicine, but there is a critical need to bring the power of AI to the patients, caregivers and clinicians who need it most,” Paul Anderson, senior vice president of research and education at Brigham and Women’s Hospital, said in the press release. “This grant will allow experts from across our state to come together to help address this key gap.”

If successful, the research would utilize AI to deliver, manage, and adapt treatment and intervention regimes for those with age afflictions. So what does that look like?


A key component, said Deepak Ganesan, a professor in UMass Amherst’s Robert and Donna Manning College of Information and Computer Sciences, will be improving on the technologies that already exist in devices like smartphones and Apple Watches.

“[We] may look at leveraging existing mobile and wearable devices such as smartphones in new ways,” he said in an email. “For example, voice-based interaction using a smartphone may be used to look at changes in the voice patterns that can be used to detect subtle changes in cognitive and physical function for patients with Alzheimers.”

Devices like Apple Watches and Fitbits, which track the steps of the wearer, can be inaccurate when used by older users, he said, because they are not calibrated to track lower speeds. And new sleep trackers can lose accuracy in users with sleep disorders or who wake up to take medications.

“Some of the focus will be on adapting the algorithms such that they can be more accurate when monitoring older adults with a range of impairments,” Ganesan said.

The center will also work with new technologies, he said, like devices that allow for patient monitoring without requiring them to wear anything.

And a key component of the research will be distilling the data gathered from patient cohorts and presenting it to patients, caregivers, and clinicians in a digestible way. Together, the adapted technologies and data could create a new system for monitoring elderly patients who want to remain home that sends help when its needed.


“It’s a difficult problem to develop AI-enhanced sensing technologies that work for people where they are,” Ganesan said in the press release. “How do you get good, useful data? How do you analyze this data and present it to the patient, caregiver and clinician? And then how can you intervene in a timely manner when a problem develops?”

Andrew Brinker can be reached at andrew.brinker@globe.com. Follow him on Twitter at @andrewnbrinker.