BCIwiki (bciwiki.com) — A preprint study shows that cross-subject cumulative learning and transfer learning can substantially reduce calibration needs for movement-related cortical potential (MRCP)-based brain-computer interfaces in children with cerebral palsy, achieving 91% accuracy without within-session calibration. The study, by R. Saadatyar, D. Damiano, and A. Behboodi, was uploaded to arXiv on July 18, 2026, and has not been peer reviewed.
The researchers collected EEG during repeated ankle dorsiflexion tasks across 27 sessions in four children with cerebral palsy and evaluated a bidirectional long short-term memory (Bi-LSTM) network. They compared seven training strategies, ranging from conventional within-session calibration to cumulative cross-subject learning and transfer learning. Cross-subject cumulative learning achieved 91% accuracy without within-session calibration, while adding transfer learning increased accuracy to 93% with minimal within-session calibration. Both approaches significantly outperformed conventional calibration strategies and achieved the highest F1-scores and receiver operating characteristic (ROC) performance, demonstrating robust generalization across sessions and participants.