Neural Mechanisms of Mixed Speech and Grasp Representation in Sensorimotor Cortices

Recent brain-machine interface (BMI) studies have challenged traditional views of functional specialization in human motor cortices, suggesting that regions associated with hand control also support speech. The extent of this dual functionality as well as the neural mechanisms underlying it are unclear. We address this by analyzing intracortical neural activity from seven brain regions (spanning motor, premotor, somatosensory and parietal regions) across two participants with tetraplegia. Across all regions, grasp decoding was robust. In addition, we achieved reliable discrete-word decoding during silent reading as well as vocalized speech. Both tasks largely recruited overlapping neural populations within each region, yet these populations reconfigured their functional connectivity between tasks. Additionally, subspace analyses revealed segregated computations for speech and grasping despite mixed selectivity at the unit level. Our findings support multi-functional BMIs capable of decoding speech and grasping from the same implant and highlight ventral premotor area 6r as a novel target.