Abstract:
Stroke is the leading cause of motor impairment worldwide. In the last decade, brain-machine interfaces (BMIs) have emerged as a promising tool for stroke rehabilitation, but the rehabilitative outcome is still moderate. This thesis capitalizes on the neuroscience behind sensorimotor integration and memory formation that occurs during stroke rehabilitation to design a novel BMI-based rehabilitation intervention that includes peripheral electrical stimulation and a sleep interface. The first study investigates the challenges and biases interpreting EEG during a visuomotor task, revealing confounding factors that might bias EEG analyses of a reaching task, a critical part of a BMI-based stroke rehabilitation. The second study builds upon the neuroscientific knowledge on memory consolidation and uses state-of-the-art techniques, like targeted memory reactivation (TMR), to manipulate memory consolidation after neuroprosthetic learning, the type of learning that takes place during a BMI training. In line with previous animal studies, increased spindle and slow wave activity was found over the brain area linked to the BMI control in the intervention night, showing that, after TMR, the occurrence of such neural correlates was more lateralized towards that brain area. The third and final study integrates peripheral electrical stimulation (PES) in a BMI system and explores how this stimulation, when properly synchronized with supraspinal activity, can excite the sensorimotor system, potentially leading to a better rehabilitative outcome. The findings of this study clearly show a higher excitatory state of the sensorimotor system, but no behavioural changes (in a motor learning task) could be found. Finally, both studies ended with a test on stroke population. Study 2 studied how sleep changes after a BMI-based rehabilitation intervention and looked for the same patterns of memory consolidation that could be found in healthy participants. Study 3, on the other hand, implemented and tested in a two-month intervention a portable home-based BMI system for rehabilitation that delivered closed-loop PES to increase excitability of the sensorimotor system. Altogether, this thesis sets the basis towards a neurorehabilitation system combining neural interfaces, peripheral stimulation, and sleep.
Brain-Machine Interfaces