Re-establishing cortico-muscular communication to enhance recovery: development of a hybrid brain-computer Interface for post-stroke motor rehabilitation

Stroke is a leading cause of adult serious and long-term disability. Notably, improving upper limb functioning is the primary therapeutic goal in stroke rehabilitation to maximize patients' functional recovery and reduce long-term disability. Nowadays, Brain-Computer Interfaces (BCIs) can be used as add-on to traditional therapies to activate rehabilitative devices directly decoding the brain activity of the user noninvasively, e.g. by means of electroencephalogram (EEG). However, the consequences of a stroke involve regions apart from the focal lesions due to disruption of connections along neural pathways. Therefore, a BCI system for motor rehabilitation should allow to train both brain and peripheral activity, reinforcing the volition that is brain control over muscular activation together with physiological muscular activation patterns. In this PhD thesis, Cortico-Muscular Coupling (CMC), which measures the synchronization between central and peripheral activation (recorded respectively through EEG and electromyogram – EMG), was studied as hybrid feature to detect movement attempts and to reinforce the physiological brain control of muscles activity. The widespread functional brain-muscle connectivity (derived from multiple EEG-EMG pairs) was characterized and compared in healthy subjects and stroke patients by means of indices derived ad-hoc from graph theory. CMC resulted to contain information about the movement type performed as well as the general clinical status of stroke patients in terms of their hand functionality, showing a high potential to be used as input of hybrid BCI (h-BCI) systems. Thus, a processing pipeline for the translation of CMC computation and the consequent CMC-based movement detection from offline to real-time was defined and optimized. A novel h-BCI prototype aimed to Re-establish Cortico-Muscular communication was developed and its feasibility was validated. Moreover, a study on the feedback delivery strategy (i.e. Functional Electrical Stimulation - FES) was performed with the ultimate aim of tailoring the stimulation to patients’ impairment. Such rehabilitative prototype recognizes close-to-normal EEG-EMG coupling during hand movement attempts, taking into account both the CMC features to reinforce during the h-BCI training, and the ones to discourage to avoid the maladaptive movement abnormalities typical of post-stroke recovery. Upon movement detection, it triggers the delivery of FES to the target muscle to support full movement execution. Such system resulted to be reliable and easy-to-use with high accuracy and timing. The developed hybrid device would allow to follow patients along recovery with a strategy tailored on their rehabilitative stage and hence maximizing the time and amount of functional recovery with potentially high impact on the stroke survivors' quality of life (personalized medicine).