Human learning of noninvasive brain-computer interfaces via manifold geometry.

Leveraging intrinsic brain activity geometry accelerates human learning of noninvasive brain-computer interfaces, with success linked to alignment with the brain's natural manifold.

• Why it matters: Current BCI adoption is hindered by slow, inconsistent learning across users, limiting their practical use in restoring or enhancing human capabilities. Understanding how to optimize learning can significantly improve BCI effectiveness and accessibility.
• What they did: Researchers used real-time fMRI to train participants to control a video game avatar by modulating brain regions involved in spatial navigation, and manipulated the mapping between brain activity and avatar movement to test the role of neural manifold constraints.
• The result: Participants successfully learned control when new mappings aligned with the brain's intrinsic activity directions, demonstrating that manifold geometry guides complex cognitive learning and offering a principle to enhance future neurotechnologies.