Moving beyond the motor cortex: A brain-wide evaluation of target locations for intracranial speech neuroprostheses

Speech brain-computer interfaces (BCIs) offer a solution for those affected by speech impairments by decoding brain activity into speech. Current neuroprosthetics focus on the motor cortex, which might not be suitable for all patient populations. We investigate potential alternative targets for a speech BCI across a brain-wide distribution. Thirty participants are recorded with intracranial electroencephalography during speech production. We continuously predict speech from a brain-wide global to a single-channel local scale, across anatomical features. We find significant speech detection accuracy in both gray and white matter, no significant difference between gyri and sulci, and limited contribution from subcortical areas. Potential targets are located within the depths of and surrounding the lateral fissure bilaterally, such as the (sub)central sulcus, the transverse temporal gyrus, the supramarginal cortex, and parts of the insula. The results highlight the potential benefit of extending beyond the motor cortical surface and reaching the sulcal depth for speech neuroprostheses.