Related Experiment Video
Updated: Jun 29, 2026
Home-Based Monitor for Gait and Activity Analysis
Published on: August 8, 2019
Real-time detection and subtyping of "On-meds" freezing of gait in Parkinson's disease using lower-limb acceleration
Wenbiao Xian1, Bin Hu2, Taylor Chomiak2
1Department of Neurology, Guangdong Provincial Key Laboratory of Diagnosis and Treatment of Major Neurological Diseases, National Key Clinical Department and Key Discipline of Neurology, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China.
Background:
Although many patients with Parkinson's disease (PD) report experiencing episodes of freezing of gait (FOG) at home under ON medication ("On-meds") conditions, objective and accurate diagnosis of different types of FOG events remains an extremely challenging task.
Methods:
We conducted an observational, case-control study enrolling 75 consecutive PD patients, who were classified into "freezer" (n = 50) and "non-freezer" (n = 25) group, based on responses to the FOG Questionnaire and clinical confirmation. A modified timed up and go (TUG) protocol comprised one single‑task TUG (sTUG) and two dual‑task TUGs (cognitive, manual). Synchronized video and a single wearable sensor (Ambulosono) provided parallel capture for real‑time detection and subtyping of FOG.
Results:
In the "freezers" group, 337 FOG episodes occurred during TUG testing, whereas none were recorded in the "non‑freezers" group (p < 0.01). The mean frequency was 2.25 FOG episodes per person per test. Compared with the sTUG test (74 FOG episodes), dual-task trials identified 263 FOG episodes, representing a 255% increase (p < 0.01). Trembling, shuffling and akinetic subtypes were identifiable on the device display (GMGI), with sensitivity 87.2% and specificity 89.5% versus video. Wearable data also localized subtypes by gait phase (initiation, turning, midway and ending).
Conclusions:
"On-meds" FOGs can be objectively diagnosed among self-reported freezers using a dual task protocol during gait tests. Parallel video and wearable sensor recordings facilitate the real-time detection and subtyping of "On-meds" FOGs, which can substantially improve the current clinical practice.
