Action sequence guidance with exposure trajectory technology improves performance of motor imagery-based brain-computer interface.

Background: The paradigms greatly influence the performance of motor imagery (MI)-based brain-computer interfaces (BCI) by guiding subjects to imagine. How to make the guidance clear and intuitive is important for MI-BCI to improve performance.

New Methods: This study proposes a novel MI-BCI paradigm based on action sequence (AS) guidance, which visualizes and choreographs sequential actions to support motor imagery. In a drawing task, the action exposure trajectory technique presents a gray nib at the starting point of the next stroke while the current stroke is being drawn, highlighting the order and details of the movement. Ten subjects participated in offline and online experiments under both AS and traditional MI conditions. EEG activation regarding multiple frequencies and periods, and MI-BCI performance are evaluated.

Results: The AS paradigm evokes more significant ERD/ERS features, and improves offline and online BCI accuracies and information transfer rates to 85.69 %, 78.77 %, and 15.60 bits/min, which are 8.37 %, 7.95 %, and 7.13 bits/min higher than the traditional paradigm. In addition, the subjects are demonstrated more comfortable subjective feelings.

Comparison With Existing Methods: The AS paradigm offers clearer and more intuitive guidance, enhances EEG feature activation, and significantly improves MI-BCI performance in both offline and online experiments.

Conclusions: Dynamic action sequences action with exposure trajectory technique could enhance the subject's brian activation by offering richer content and more intuitive guidance, providing a new way for prompting BCI performance.