Rhythm-Fluctuation-Based Evaluation Platform For Gait Training Of Parkinson'S Disease Patients

Gait training using rhythmic auditory cue showed training effect on gait dynamics of patients with Parkinson's disease (PD). On the other hand, 2 indicators of gait rhythm fluctuation, coefficient of variation (CV) and scaling exponent a, can evaluate severity of PD. However the gait training was not evaluated by both of these indicators at a time. In this paper, we aim to propose rhythm-fluctuation-based evaluation platform for gait training of PD patients. This platform consists of CV and a of stride interval. Specifically, we evaluated 3 types of gait training by calculating the change amount of CV and a. The training types are 1) interactive Walk-Mate (WM) gait training, 2) fixed-tempo Rhythmic Auditory Stimulation (RAS) gait training, and 3) Silent Control gait training. The training program was continued for 4 days. The first trial in each day was baseline walking trial without rhythmic cue. The second and the third trials were training trial corresponding to the training condition. Twenty-four PD patients walked for about 2 minutes in each trial. These patients divided into 3 groups about training type. As a result, WM gait training improved alpha and CV significantly. RAS gait training tended to improve CV, but it seemed to worsen alpha of gait rhythm. These effects of RAS gait training were marginally significant. However, Silent Control training did not show the significant effect on both of the indicators. From these results, the rhythm-fluctuation-based evaluation platform was successfully used to evaluate each type of gait training. In addition, the rhythm-fluctuation-based evaluation platform detected the difference between WM gait training and RAS gait training by alpha.