Classifying Gait Disorder in Neurodegenerative Disorders Among Older Adults Using Machine Learning

Gait disorders are a significant concern for older adults, particularly those with neurodegenerative diseases such as Parkinson’s disease, Hunting-ton’s disease, and Amyotrophic Lateral Sclerosis. Accurately classifying these conditions using gait data remains a complex challenge, espe-cially in older populations, due to age-related changes in gait patterns, comorbidities, and increased variability in mobility, which can obscure disease-specific characteristics. This study explicitly classifies neurode-generative diseases in older adults by analysing age-specific gait force data. Continuous Wavelet Transform (CWT) was utilised for advanced feature extraction, capturing both temporal and spectral signal character-istics. Classifiers including Support Vector Machine (SVM), Random Forest (RF), Decision Tree (DT), and Multilayer Perceptron (MLP) were em-ployed. The results demonstrated that SVM achieved an accuracy of 87.5%, outperforming RF and MLP, which achieved 83.3% and 50.0%, respec-tively. These findings underscore the importance of using tailored machine learning approaches to improve the diagnosis and management of neurodegenerative diseases in older adults. The potential for real-world application includes integration into clinical settings, enabling early detection and personalized interventions for individuals with gait disorders.