Spatiotemporal encoding signatures of covert speech revealed by source-localized EEG and fMRI
Year: 2026
Type: article
Abstract: Covert speech, the silent production of words without overt articulation, is fundamental to human cognition and a potential target for restoring communication in speech-impaired populations. Yet where, when, and how the human brain represents specific phrase content during covert speech remains largely unresolved. Here, we combine source-localized EEG (n = 57) with fMRI validation in the same cohort (n = 49) to map the encoding of five covert command phrases. Phrase-discriminative activity is concentrated before estimated articulation onset, peaking at 200–400 ms and strongest in the theta band. Cross-modal representational similarity analysis reveals convergent EEG–fMRI phrase representations in left sensorimotor cortex, while phase-lag connectivity identifies left postcentral gyrus as a phrase-sensitive functional hub within a parietal-somatosensory-temporal sequence during the speech planning period. Beyond these group-level signatures, individual-specific encoding patterns substantially exceed group-average effects. These findings provide a temporally precise, cross-validated cortical map of covert speech and identify targets for speech neuroscience and non-invasive brain-computer interfaces. Source-localized EEG with fMRI validation in the same cohort maps where and when the cortex encodes covert speech. Phrase information peaks 200-400 ms after the cue in left sensorimotor cortex, and individual-specific patterns exceed group effects.
Source: Communications Biology
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $4,190
OpenAlex ID: W7220731861
