University of Delaware Opens AI Institute to Help Personalize Rehabilitation
The five-university effort will combine patient assessments, medical records and movement data, with researchers emphasizing support for clinicians rather than replacement.
Loading page…
The five-university effort will combine patient assessments, medical records and movement data, with researchers emphasizing support for clinicians rather than replacement.
Listen to this story
The institute will study whether combining clinician assessments with medical records and ongoing movement data can support more tailored rehabilitation for people with mobility-related conditions. The five-year effort is part of the National Science Foundation’s AI institute program, its 30th such center; reported funding differs slightly, at $21 million or $21.5 million. Researchers also emphasize building and controlling their own systems, rather than relying on third-party AI vendors—a practical constraint on how the work is developed and tested.
The collaboration includes Delaware State, Princeton, the University of Pennsylvania, and the University of Florida.
The initial research focus is rehabilitation for stroke, Parkinson’s disease, and other mobility-related conditions.
Researchers envision using data from devices such as an Apple Watch alongside medical histories and in-person assessments.
More personalized rehabilitation is the ambition; supporting clinicians, not replacing them, is the stated design. The University of Delaware inaugurated a National Science Foundation AI institute this week for a five-year effort costing roughly $21 million. Its researchers aim to turn movement data and medical histories into tools that help providers tailor patient care.
University leaders cut the ribbon at the institute, headquartered within Delaware’s Data Science Institute, according to 6abc. Delaware leads the collaboration with Delaware State University, Princeton University, the University of Pennsylvania and the University of Florida. The initial focus is physical rehabilitation, including care for people with stroke, Parkinson’s disease and other mobility-related conditions.
The two local outlets give different funding totals: 6abc describes a $21.5 million initiative, while NBC10 puts it at $21 million. Both describe a five-year program. The center is the 30th AI institute established under the NSF program.
The proposed approach combines what clinicians learn during in-person assessments with a patient’s medical history and new information from devices such as an Apple Watch. AI would analyze that larger body of information alongside the clinician’s expertise. Movement analysis and medical records are intended to help providers choose more personalized treatment for mobility-related conditions.
Austin Brockmeier, the institute’s deputy director and co-principal investigator, described the goal as having AI perform analysis behind the scenes when a physical therapist has limited time with each patient. With more patients’ data and more frequent updates, the research approach also aims to identify patterns in which treatments and medications work best.
We want to use ai as a companion to take that expertise they have and couple it with the ability to compute on a large amount of data that has been recorded about the person
Austin Brockmeier, University of Delaware College of Engineering, speaking to NBC10
The facilities place clinical physical therapy and research physical therapy labs next to each other. Brockmeier told 6abc that patients can move between the two. Across the hallway are additional labs for developing and testing prototypes, including treadmills designed to engage people’s walking in new ways. Those spaces give the effort a setting for studying technology alongside rehabilitation care.
The institute’s human-centered focus has a practical target: helping people stay mobile and return to work, sports and everyday activities. Researchers describe more efficient treatment as a goal of the collaboration.
Control over the technology is another stated priority. John Huffman of University of Delaware research computing told NBC10 that researchers need to control these systems rather than depend on third-party vendors selling AI. He called that independence critically important to how the team conducts its research—a separate concern from how much patient information an AI can analyze.
Loading discussion...
Join the conversation
Explain where independence would matter most, or where vendor tools would be enough.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.