Frontiers in Human Neuroscience
Graded multisensory neurofeedback via the iNems system for comprehensive upper limb recovery after stroke: a multiple baseline study
By Takayuki Kodama, Ken Kitai, Yasuhisa Mizomoto and 17 others.
Abstract
Introduction Post-stroke upper-limb recovery remains challenging, especially in the presence of sensory impairments. Conventional brain–computer interfaces (BCIs) are constrained by binary feedback, single-modality stimulation, and exclusion of patients with sensory deficits. We developed the Imagery Neurofeedback-based Multi-sensory System (iNems), a neurofeedback system that provides graded multisensory (visual and electrical) feedback based on ERD magnitude. This study validated iNems’ efficacy in motor function, self-reported numbness, and cognitive functions using a multiple-baseline design (MBD). Methods Four participants with post-stroke upper-limb motor and sensory impairments underwent a multiple baseline design. The baseline phase (phase A) used random feedback; the intervention (phase B) used adaptive iNems. iNems quantified motor intention via sensorimotor EEG event-related desynchronization (ERD) and delivered graded visual and electrical feedback. Primary outcomes were the Fugl-Meyer Assessment-Upper Extremity (FMA-UE) and Box and Block Test (BBT). Secondary outcomes included body cognition, numbness, and neurophysiological (ERD/ERS) changes. Results All participants demonstrated significant improvement during the intervention. FMA-UE increased by 6–16 points. The BBT showed large effects (integrated Tau-U = 0.75). Body cognition improved significantly in both sense of agency (Tau-U = 0.80) and ownership (Tau-U = 0.89). Numbness was reduced in affected cases (Tau-U = −0.85). Neurophysiological changes correlated with functional gains, with some patterns suggesting increased neural efficiency. Conclusion This first MBD-validated demonstration shows that graded, multisensory graded neurofeedback based on ERD magnitude can improve motor, self-reported numbness, and cognitive domains after stroke. These findings support adaptive closed-loop therapy as a promising, multifaceted rehabilitation strategy.



