Preliminary research to be presented at the American Stroke Association's International Stroke Conference 2026 indicates that a smartwatch application measuring social interactions among hospitalized stroke survivors could enable new treatments to preserve cognition, social engagement, and quality of life. The app, called SocialBit, uses a machine learning algorithm to detect social interactions through acoustic patterns in the environment, specifically identifying human speech. This technology was found to be 94% as accurate as human observers in recognizing social interactions, maintaining 93% accuracy even in patients with aphasia, a language disorder common after stroke.
Researchers recruited 153 adults hospitalized for ischemic stroke, who wore an Android smartwatch equipped with SocialBit in their rooms between 9 a.m. and 5 p.m. daily for up to eight days. The app logged socialization time based on sounds from the participant or another person talking. Simultaneously, research team members watched livestream video to log minute-by-minute social interactions for comparison. The study found SocialBit's performance remained consistent despite background noise like television, different environments such as rehabilitation units versus hospitals, and across various smartwatch models.
Social interaction is known to support brain health and recovery after neurological injury. According to the American Stroke Association, loss or change in speech and language profoundly alters the social life of stroke survivors, yet socializing is one of the best ways to maximize recovery. Study lead author Amar Dhand, M.D., D.Phil., an associate professor of neurology at Mass General Brigham, noted that his previous research demonstrated stroke survivors who are socially isolated have worse physical outcomes at three and six months post-stroke. "We created a tracker of social life customized for stroke survivors. Tracking human engagement is crucial, and social isolation can now be identified in real-world situations," Dhand said. This could allow notifications to patients, family, caregivers, and health professionals about social isolation.
The research also revealed that participants with more severe strokes had less social interaction, with about a 1% drop in total social interaction minutes for each 1-point increase on the NIH Stroke Scale, a standardized severity assessment tool. Dhand expressed surprise at how well the app performed for people with aphasia, noting that using sounds instead of words to protect privacy ended up being helpful for those with limited language skills. He suggested SocialBit could support therapies like speech, occupational, and exercise therapy, and future research could use it to measure social isolation risk in hospitals and after discharge, exploring links to depression and other mental health changes post-stroke. The app might also be tested for other brain injuries and healthy aging to maintain brain health over time.
Cheryl Bushnell, M.D., M.H.S., FAHA, chair of the American Heart Association Stroke Council and chair of the writing group for the Association's 2024 Guideline for the Primary Prevention of Stroke, commented on the research, calling it "fascinating" for capturing social interactions. She noted potential applications in measuring quality of hospital care and social interactions at rehab facilities and nursing homes, though she raised questions about whether the app distinguishes between hospital personnel and non-hospital visitors. Bushnell, who was not involved in the study, is a professor and director of the Center for Transformative Stroke Care at Wake Forest University School of Medicine.
The study involved participants with an average age of 66, ranging from 26 to 94, with 53% men, 46% women, and 1% other. Strokes were mild, with a median NIH Stroke Scale score of 2, and cognitive impairment was also mild. Participants were hospitalized at Brigham and Women’s Hospital between June 2021 and March 2025, with some transferred to Spaulding Rehabilitation Center during assessment. Exclusions included those receiving end-of-life care, prior dementia diagnosis, or inability to speak or understand English well enough for surveys or consent. It is important to note that the study is a research abstract presented at a scientific meeting; findings are considered preliminary until published as a full manuscript in a peer-reviewed journal. According to the American Heart Association’s 2026 Heart Disease and Stroke Statistics, stroke is now the fourth leading cause of death in the U.S., highlighting the need for innovative recovery tools like SocialBit.
