Presentation
Multimodal Cues for Motor Imagery Guidance: Not Necessarily an Optimal Choice
DescriptionMotor-Imagery Brain-Computer Interface (MI-BCI) guidance prompts design is crucial for enhancing neural activity during imagery practices in upper limb rehabilitation and human-machine collaboration. However, the comparative effectiveness of different modality-based prompts in improving motor imagery (MI) ability remains under-explored. This study analyzed EEG signals from 30 participants performing ball-grasping tasks under four guidance conditions: static image, synchronized visual, synchronized visual-auditory, and synchronized visual-vibration. Results show synchronized visual prompts (alone or combined with other modalities) outperform static images in enhancing EEG spatial feature discrimination, ERD/ERS discriminability, system recognition accuracy, and intention maintenance. Notably, multimodal guidance does not always improve performance and may weaken neural response stability, indicating it is not necessarily optimal. Synchronized visual prompts’ immersion effectively boosts MI performance and EEG discriminability, providing insights for optimizing MI-BCI prompt design.
Contributors
Event Type
Lecture
TimeThursday, October 22nd3pm - 3:20pm PDT
Location
Similar Presentations