Neuroimaging biomarkers toward an optimized and personalized AOT

Stroke is the second leading cause of death in industrialized countries. In recent years, some scientific studies have focused on the search for new non-invasive strategies which contribute to the functional recovery of stroke patients. Action Observation Therapy (AOT) is one of them. This rehabilitative approach is aimed at facilitating neural plasticity toward the reconstruction of impaired motor functions through the stimulation of motor cortex via visual stimuli containing actions. This study was carried out fitting into this research context, proposing an analysis of the electroencephalographic signals acquired during an action observation protocol. The aim of the work is to identify, through the study of the desynchronization and synchronization index (ERD/ERS), the categories of gestures that cause greater cortical activation to optimize intervention based on AOT protocol. The results showed that the behavior of motor cortex during the visualization of motor gestures is statistically different respect to control videos (not human actions) with a p-value < 0.03 for mu band (8-13 Hz) and p-value < 0.0001 for beta band (15-20 Hz). Another key result based on time course of signals power showed that the control stimuli cause less desynchronization in the motor area compared to visual motor stimuli.