Deep Learning based Gait Analysis for Early Detection and Progression Monitoring of Alzheimer's Disease

Neurological diseases present a considerable impact on individuals by affecting their quality of life leading to disability and mortality. Gait represents the pattern of human walking, which serves as a chief indicator of health status, functional impairment, and treatment prognosis. Gait analysis (GA) plays an essential part in the assessment of neurological disorders, with patterns helping as reliable factors of potential disorders in the future. Alzheimer's disease (AD) adjacent profound concerns across universal healthcare networks demanding timely monitoring and suitable intervention. In this analysis, we present an innovative approach to model the time-based dependence in AD progression by integrating gait inspection with cognitive performance metrics and functional neuroimaging using recurrent neural networks (RNNs). By encompassing LSTM, the longitudinal nature of AD data allows movement patterns to be utilized as a supplemental marker to capture subtle changes in cognitive function as well as mobility over time. By inspect consecutive data gathered from individuals at risk or diagnosed with AD. Our approach aims to forecast future cognitive decline, with biological markers indicative of disease progression helping in early diagnosis. With accuracy, recall as 0.98, precision, F1-Score and AUC-ROC as 0.99 our integrated framework makes use of an indigenous dataset to offer a holistic understanding of the multifaceted dynamics in AD progression, paving the way for personalized care and treatment strategies tailored to suit individual cognitive and motor impairments.