A robust, cost-effective and non-invasive computer-aided method for diagnosis three types of neurodegenerative diseases with gait signal analysis

Abstract One of the challenges of computer-aided diagnostic systems is to propose a reliable algorithm detecting different types of neurodegenerative diseases using cost-effective procedures. To tackle the challenge, this study developed a new methodology based on statistical and entropic features of vertical ground reaction forces of gait and sparse coding classification technique. The effect of individual differences on the proposed and standard machine learning methods was also explored with emphasize on the severity and duration of diseases as well as the right and left foot parameters. This method was evaluated using a publicly available dataset, which contains 16 healthy control subjects, 13 patients with Amyotrophic lateral sclerosis (ALS), 15 patients with Parkinson’s disease (PD), and 20 patients with Huntington’s disease (HD). It achieved the best average accuracy rates of 100%, 99.78%, and 99.90% for ALS, PD, and HD detection, respectively. The results confirmed that the proposed algorithm can identify all diseases at both early and advanced stages using either left or right foot features.