Neurodegenerative diseases (NDDs) such as Parkinson’s disease (PD) and stroke often manifest as altered gait rhythms. While various clinical and progression metrics are available to evaluate gait dynamics, the underlying sensorimotor mapping between normal and abnormal gait patterns remains insufficiently explored. These gait patterns are inherently non-stationary, non-linear, and non-parametric, posing significant challenges in their characterisation. Spatio-temporal to spatio-spectral features were essential for capturing the intricate complexities of impaired gait dynamics, as reflected in fractal dimensional components and entropy-based nonlinear feature space mappings. This systematic review chapter aims to bridge the existing knowledge gaps by providing comprehensive insights into the sensorimotor alterations observed in NDDs. Recent feature engineering and gait analysis advancements offer a detailed understanding of how frequency-domain and spectral-level insights can enhance disease detection and monitoring. Additionally, the findings presented here contribute to the development of non-invasive neuromodulation techniques, which hold promise in restoring motor function and improving the quality of life of patients. Overall, this review is a valuable resource for clinicians, researchers, and engineers seeking to explore the sensorimotor mechanisms underlying gait abnormalities further. It also underscores the importance of applying robust machine learning (ML) algorithms and feature selection approaches to achieve more accurate and reliable assessments of NDD progression.