Decoding speech intent from non-frontal cortical areas
Date:
Brain–computer interfaces (BCIs) for speech restoration have largely focused on neural activity recorded from frontal speech motor regions, including ventral precentral cortex, postcentral gyrus, and inferior frontal gyrus. In contrast, temporal and parietal cortices are traditionally associated with speech and language perception, and their role in speech production remains unclear.
In this talk, I present evidence that temporal and parietal cortical areas carry a decodable signal related to speech intent during speech production. Using electrocorticography (ECoG) and stereoencephalography (sEEG) recordings from human participants, we show that speech intent can be reliably detected from these regions prior to or during speech output.
Identifying intent-related signals is critical both for practical speech BCI deployment and for addressing ethical concerns related to unintended decoding of internal speech or thought. These results suggest that non-frontal cortical areas may provide complementary signals for speech BCIs, particularly in individuals with frontal lobe damage such as aphasia.