BCIwiki (bciwiki.com) — A sensory-guided joint learning framework achieved average online discrete accuracies of 86.0% for 1D and 77.5% for 2D motor imagery tasks in 31 BCI-naïve participants, researchers at Carnegie Mellon University reported in Nature Communications on July 15, 2026. The framework integrates human motor learning with adaptive machine learning to improve BCI training and performance.
Tactile guidance reduced user exploration and accelerated neural adaptation, while sample reweighting aligned decoder updates with human learning trajectories. Continuous control accuracies were 77.5% (1D) and 66.9% (2D). The authors suggest this approach advances BCI training from passive calibration to active human-machine joint learning, with potential applications in communication and rehabilitation.