Gait Rhythm Fluctuations Assessment for Neurodegenerative Patients

Abstract

Neurodegenerative diseases (NDDs) such as Parkinson’s disease (PD), Amyotrophic Lateral Sclerosis (ALS) and Huntington Disease (HD) are identified as the deterioration of motor neurons in human brain. These diseases would manipulate the strides from one gait cycle to another. Therefore, gait assessment can yield a significant approach for synthesizing a noninvasive technique to evaluate the effects of the neurological morbidness on the human gait dynamics and its variation with diseases. The present study explores the improvement of the classification capability by using nonlinear features with previously used linear features. Fisher score (FS) selection strategy has been used to get the optimal feature subset and the optimal gait time series for classifying NDD. Support vector machine (SVM) with radial basis kernel function (RBF) has been used to classify NDD patients against healthy control (CO) ones with an overall accuracy 95.31%.