Researchers at Pitt’s School of Health and Rehabilitation Sciences are developing AI to help train therapists to administer better post-stroke rehab.
Researchers from the University of Pittsburgh School of Health and Rehabilitation Sciences are developing AI to help improve how therapists administer post-stroke rehabilitation.
Elizabeth Skidmore, a rehabilitation scientist, and professor in the Department of Occupational Therapy, is a proponent of a rehabilitation method called strategy training, which shifts a rehabilitation therapist’s role from an authoritative instructor to one of a supporting player in a patient-driven process.
Skidmore’s training method purports to give stroke survivors more control over their recovery by training them to prioritize the tasks that matter to them, develop a plan to execute the activities, and practice problem-solving skills.
Skidmore has conducted three randomized, controlled clinical trials to examine the effectiveness of strategy training and aspires to run a national multisite trial of strategy training in rehabilitation facilities. To make Skidmore’s training more replicable, she is exploring using AI that experienced therapists have trained.
Researchers hope to train AI using transcripts from video-recorded rehabilitation sessions and test its accuracy against a trained human fidelity rater.
To help turn observations into data, Skidmore employs fidelity raters — licensed occupational therapists and occupational graduate students — to watch recorded rehabilitation sessions and complete a checklist as the clinician uses appropriate cueing strategies. Examples include asking open-ended questions and using guiding statements rather than direct skill training.
“On average, the therapists we’ve evaluated are using guided cues 5% of the time,” said Skidmore. “Our studies suggest increasing guided cues to 40% or 50% of the time can significantly improve client outcomes. It just requires training therapists to monitor and change their habits.”
Researchers noted their progress in a paper to be published at the AMIA 2023 Informatics Summit. Hunter Osterhoudt, a graduate student in the Department of Computer Science, is the first author of the paper and will present this work at the conference.
So far, the AI’s verbal processing results have met industry reliability standards and responded to the challenges inherent in Skidmore’s project.
Looking ahead, the team plans to integrate computer vision next, training the algorithm to recognize different types of physical gestures used in the rehabilitation procedure.
Skidmore estimates the automation project will be in development for a few years before it’s ready and available for commercialization and widespread use.