Recognition of Neurodegenerative Diseases with Gait Patterns using Double Feature Extraction Methods

Neuro Degenerative Diseases (NDD) are instigated due to the damage caused in the human brain neurons. This will cause changes in the walking pattern of the disease affected persons. Human gait analysis is one of the best tools to identify every person by their unique walking pattern. The studies state that lacks in clinical diagnosis to identify the neurodegenerative disease is making the person's health condition worse. The accurate diagnosis of brain disorder in a simplified and machine intelligence way will help the doctors in the early treatment progression. This paper is presented to classify and recognize neurodegenerative disease (NDD) patients with gait patterns using double feature extraction methods such as Recurrence Quantification Analysis (RQA) and Fast Walsh-Hadamard Transform (FWHT). To recognize the neurodegenerative diseases such as Parkinson’s, Amyotrophic Lateral Sclerosis and Huntington’s Diseases Machine learning techniques were applied. RQA is applied to quantify the input gait left swing pattern and right swing pattern and FWHT is utilized to extract the statistical features. These RQA and FWHT features extracted are further used for recognition of neurodegenerative diseases using Discriminant Analysis, Random forest, Multi SVM and KNN classifiers gives the result of 92.19%, 89.84%, 89.06%, and 89.84% accuracy respectively. Therefore, it is possible to classify and recognize neurodegenerative disease patients by the machine learning algorithms with the gait patterns.