Real-time detection and subtyping of “On-meds” freezing of gait in Parkinson’s disease using lower-limb acceleration data

Key Points

• Freezing of gait was accurately detected using lower-limb acceleration data, improving management strategies for Parkinson's disease.
• The method achieved a high accuracy rate with a sensitivity of over 85% in detecting freezing episodes.
• Analysis of lower-limb acceleration data allowed for effective subtyping of freezing of gait, revealing patterns in symptoms.
• May enable clinicians to tailor treatments based on specific freezing of gait characteristics using real-time data.