News release

From: The Royal Society

Freeze warning- A personalised AI tool may be able to predict and warn Parkinson’s disease patients of gait freezes, a disabling symptom of the disease. The tool, combined with a wearable ankle sensor, was tested with six patients at different stages of their daily medication cycle. It was trained on the specific walking style of each individual and was able to predict freezes with nearly 99% accuracy which could help reduce falls and freeze-induced anxiety. Royal Society Open Science

Personalized Prediction System for Early Prediction of Freezing of Gait in Parkinson's Disease Using Explainable AI

Royal Society Open Science

Freezing of gait is a common and disabling symptom of Parkinson’s disease, where people suddenly feel unable to move their feet, increasing the risk of falls. This study presents a personalized artificial intelligence approach that predicts freezing episodes before they occur. The system learns each patient’s unique movement patterns rather than using a single model for everyone. Validated on six patients, it achieved nearly 99% accuracy and worked fast enough for real-time use. Importantly, the system explains its decisions, showing which movement signals matter most. This approach could help improve safety, support timely interventions, and enhance quality of life for people with Parkinson’s disease.