Motion analysis of real-time somatosensory cue on freezing of gait during turning in people with Parkinson's disease

Abstract

Background: Freezing of gait (FOG) has a substantial impact on balance and walking in people with Parkinson’s disease (PD). Previous studies have shown that insufficient weight shift to unload the swinging leg may contribute to gait discontinuation. Purpose: This study aimed to examine the immediate effect of a real-time somatosensory cue to facilitate lateral weight shift during FOG in people with PD. Methods:Wedesigned awearable cue devicewhich composedof a sensor part and stimulator part. The sensor part is an inertial sensor placed at shank, to detect FOG episodes with the embedded time-frequency analysis algorithm. The stimulator part is a vibrator pad placed below the lateral malleolus, to facilitate lateral weight shift during FOG. Eighteen idiopathic people with PD (Age: 68.4± 4.2 year) having FOG were recruited, and took this study during their medication OFFperiod. Participants conducted the timedUp-and-Go test with the stimulator part turned on (cued trials) and off (nocue trials). Participants were videotaped and the clips were reviewed by two physical therapists to identify the freezing episodes. Weight shift was evaluated by the inclination angle of the joint line of center of mass and center of pressure. FOG severity was evaluated by FOG number and FOG duration during turning. Turning performance was evaluated by turning time and turning steps. Results: The sensor part showed moderate-to-high accuracy in detecting FOG episodes with respect to therapists’ visual inspection (sensitivity: 0.78; specificity: 0.88). Compared with no-cue trials, participants in the cued trials had greater frontal inclination angle during FOG (cued: 8.17± 0.98 vs. no-cue: 7.26± 0.53, degree). In addition, the cued trials had decreased FOG duration (cued: 2.12 vs. nocue: 3.62, second) and turning time (cued: 5.49± 2.48 vs. no-cue: 8.67± 3.70, second) than the no-cue trials, despite no difference in turning steps (cued: 10.6± 4.2 vs. no-cue: 12.2± 5.0, steps) and FOG number (cued: 0.8 vs. no-cue: 1.1, FOG per turning). Conclusion(s): Our results showed that the real-time somatosensory cue could help gait reinitiation by facilitating lateral weight shift during FOG. The shortened FOGduration decreased the turning time in people with PD. Implications: A wearable device with inertial sensor is capable of detecting FOG episodes and controlling realtime somatosensory cue to facilitate weight shift. This could provide a potential intervention for FOG during daily activities in people with PD.