A Closed-Loop Tactile Stimulation Training Protocol for Motor Imagery-Based BCI: Boosting BCI Performance for BCI-Deficiency Users.

Abstract

Background: Brain-computer interfaces (BCIs) enable users to control and communicate with the external environment. However, a significant challenge in BCI research is the occurrence of "BCI-illiteracy" or "BCI-deficiency", where a notable percentage of users (estimated at 15 to 30%) are unable to achieve successful BCI control. For those users, they are struggling to generate stable and distinguishable brain activity patterns, which are essential for BCI control. Existing neurofeedback training protocols, often rely on the trial-and-error process, which is time-consuming and inefficient, particularly for these low-performing users.

Methods: To address this issue, we propose a closed-loop tactile stimulation training protocol, in which tactile stimulation training is incorporated within the closed neurofeedback loop, providing users with explicit guidance on how to correctly perform MI tasks. When a subject performs an incorrect MI trial, tactile-assisted MI training is provided to guide the user toward the correct brain state, while no training is given during correct performance.

Results: The results from our study demonstrated that the proposed training protocol significantly enhances BCI decoding performance, with an improvement of 16.9%. Moreover, the BCI-deficiency rate was reduced by 61.5%. Further analysis revealed that the training process also led to enhanced motor imagery-related cortical activation.

Conclusion: The proposed training protocol significantly improved BCI decoding performance, enabling previously BCI-deficient users to surpass the 70% control threshold.

Significance: This study demonstrates the effectiveness of closed-loop tactile-assisted training in enhancing BCI accessibility and efficiency, paving the way for more inclusive neurofeedback-based BCI training strategies.