Abstract: Objective To observe the effects of visual and audio intelligence system on freezing of gait(FOG) in the patients with Parkinson's disease(PD). Methods Based on a wearable inertial sensor equipment, a set of automatic detection and intelligent intervention system for FOG of PD was established to observe the changes of FOG through three different intervention methods including visual, audio and the combination of audio-visual. Furthermore, visual intervention with different distances and audio intervention with different step frequencies were used to intervene the specific patients. Results The wearable inertial sensor device system can detect most FOG events in real time, and the detection rate was 96.67%. The results of different intervention methods showed that compared with the non-intervention group, the FOG frequency was decreased, the maximum freezing index was reduced, and the average step length was increased after the visual intervention. After audio intervention, the FOG frequency was decreased and the turn time was shortened. After the combined intervention, the number of FOG frequency and the average duration of FOG were significantly decreased, the maximum and average freezing index were significantly reduced, and the turn time, average stride frequency and average step length were increased. Six patients showing more significant effects after visual intervention were further intervened with different laser line spacing. The results showed that the FOG frequency was decreased, average freezing index was reduced, and the step length was increased in the patients with 1-time step length compared with those with 1.5-time step length. The freezing evaluation indexes and gait evaluation indexes in the patients with 1-time step length were both improved compared with those with 2-time step length. Six patients showing more significant effects after audio intervention were further intervened with different audio step frequencies. Compared with the patients with 0.9-time audio step frequency, the number of FOG frequency and FOG duration were decreased, the average and maximum freezing index were reduced, the turn time was shortened, the average stride frequency and step length were increased in the patients with 1.1-time audio step frequency. Compared with the patients with 1-time audio step frequency, the number of FOG frequency and average freezing index were decreased, the average stride frequency and step length were increased in the patients with 1.1-time step frequency. Conclusions The combination of visual and audio intervention shows more significant relief of FOG, and the intelligent intervention could significantly shorten the duration of FOG, and improve the turn speed, stride frequency and stride length of patients.

Key words: Parkinson's disease, freezing of gait, wearable devices