Wearable knee acoustics
Listening to the sounds joints make as they move, with wearable sensors, to assess joint health outside the clinic.
Juvenile idiopathic arthritis (JIA) is the most common chronic condition in pediatric rheumatology, yet disease activity is still judged mainly through periodic clinical exams and imaging. Knees produce acoustic emissions as they move, and these change with joint health. I work on turning those signals, recorded by a small wearable, into objective measures of disease activity that could be tracked between clinic visits.
- Clinical validation. I validated KneeMS, a low-cost wearable knee-acoustic sensing system, against a benchtop reference in 36 children with JIA, finding strong agreement on key acoustic features (Spearman’s ρ up to 0.79).
- Disease-activity prediction. I developed machine learning pipelines that classify active versus inactive JIA knees from wearable acoustic signals, with prediction probabilities that track clinical severity.
- Toward home monitoring. Exploratory longitudinal results followed clinically expected changes across follow-up visits, and most children reported no discomfort wearing the device.
Related papers:
- KneeMS to monitor knee acoustic emissions for juvenile idiopathic arthritis, IEEE JTEHM 2026
- Predicting disease activity in juvenile idiopathic arthritis using KneeMS and machine learning, IEEE EMBC 2026 (in press)