Gait Analysis for Fall Prediction using EMG-triggered Movement-Related Potentials

Abnormal gait is an usual feature in neurodegenerative disease (i.e.: Huntington Chorea, Parkinson and Alzheimer), while the capability to maintain a stable posture and fluid walking is progressive impaired in aging. Monitoring and correcting the insurgence of abnormal dynamic balance opens new scenarios in the cure of these diseases and falls prevention. In this work, we present a study based on EEG time-frequency analysis to identify the correlation between synchronized EEG and EMG signals for gait analysis. Several tools for gait analysis are developed and experimented i.e. EMG trigger generation with dynamic threshold, EMG co-contraction, EEG movement related potentials (MRPs) and EEG event related desynchronizations (ERDs). This work particularly focus on gait analysis indexes implementation and experimentally obtained results based on a large dataset, including different type of gait i.e. normal gait, perturbed gait and gait during a second cognitive task (DT). A weighted average on the calculated indexes are exploited to quantify the falling risk