Scaling Cues to Freeze Gait: An Augmented Reality Approach to Benchmark Freezing of Gait in Parkinson’s Disease
Scaling Cues to Freeze Gait: An Augmented Reality Approach to Benchmark Freezing of Gait in Parkinson’s Disease
Where did the research take place?
The study site has not been established. Author addresses may differ from where the research occurred.
Publication status: preprint
A plain-language reading has not been prepared for this paper yet.
Original abstract
Background Freezing of gait (FOG) affects up to 80% of people with advanced Parkinson’s disease and is difficult to elicit reliably during clinical assessments. Augmented reality (AR) offers potential for standardized FOG provocation by presenting virtual triggers in any environment. Objective To evaluate whether an AR-based turning task could elicit FOG in a graded, dose-dependent manner and assess user experience with the technology. Methods Thirteen people with Parkinson’s disease (8 freezers, 5 non-freezers) completed an AR pillar-turning protocol across two cohorts: clinic-based (n=4, all freezers) and laboratory-based (n=9, mixed). Participants performed 360° turns around a virtual pillar presented at three diameters (0.6, 0.4, 0.2 m) using a Microsoft HoloLens 2, manipulating turning radius to vary task difficulty. FOG episodes were video-recorded and independently annotated. Participants completed perception questionnaires and the New Freezing of Gait Questionnaire (NFOG-Q). Results All clinic freezers exhibited FOG during AR pillar turns, with a clear dose-response relationship: mean episodes increased from 2.3 at 0.6 m to 5.3 at 0.4 m to 8.5 at 0.2 m diameter. No laboratory participants experienced FOG during pillar turns, though one lab freezer froze during return turns. NFOG-Q profiles indicated comparable daily-life FOG severity between clinic and laboratory freezers, suggesting environmental factors drove differential outcomes. Participants reported positive experiences with AR quality, safety, naturalness of movement, and rapid adaptation, though clinic participants reported higher immersion than laboratory participants. Conclusions AR-based pillar-turning successfully elicited graded FOG in susceptible individuals within a FOG-provoking environment, demonstrating proof-of-concept for scalable virtual trigger paradigms. Effectiveness depends on matching environmental context to individual FOG susceptibility, with implications for standardized clinical FOG assessment.