Multimodal EEG and fNIRS dataset for motor imagery (upper-limb joints)

• Open multimodal dataset of simultaneous EEG and fNIRS for motor imagery of eight unilateral upper-limb joint tasks from 18 subjects (5,760 trials total; Scientific Data, 2025).¹
• Tasks span hand, wrist, elbow, and shoulder actions (open/close, flexion/extension, abduction/adduction, pronation/supination); validity checked via activation patterns and classification performance.¹
• Supports hybrid EEG–fNIRS decoding benchmarks beyond binary left/right MI paradigms.¹
• Raw data on figshare (DOI 10.6084/m9.figshare.24123303) with MATLAB/Python loading and decoding code in FJMI_EEG_fNIRS.²

Footnotes

1. https://doi.org/10.1038/s41597-025-05286-0 ↩ ↩² ↩³

2. https://github.com/Bill-Chen-BJ/FJMI_EEG_fNIRS ↩