Learning-induced Neural Plasticity in the Primary Motor Cortex during the Motor Imagery Task and the Speech Task

Abstract

Brain-computer interfaces (BCIs) are systems that provide the connection between the brain and the external device. With the development of BCIs, understanding the basic mechanism like the BCI-induced brain plasticity enables the practical design of training sessions. We recorded neural activity over five weeks during the motor imagery task and the speech task under our spike-based BCI system involving an individual with tetraplegia. Pairwise correlations between neurons were calculated to quantify brain plasticity for the neural ensemble. The sharply increasing correlated activity was observed at the initial training phase and then declined and remained relatively low for both tasks. Our results also demonstrate the BCI-induced neural plasticity during short-time training sessions for the motor imagery task and the speech task. Moreover, a stronger correlation was found during the speech task than during the motor imagery task, suggesting task-induced changes in neural connectivity. Our findings provide important insights into human brain plasticity and thus benefit the design of clinical neuroprosthetics systems.