Two-component spatiotemporal template for activation-inhibition of speech in ECoG

TL;DR

The paper addresses how speech is represented in the sensorimotor cortex at high spatial resolution by examining cross-frequency interactions between $\Gamma$ and $\beta$ bands in ECoG data. Using PCA to compress 256-channel bandpowers into a two-component (plus a small third component) spatiotemporal framework and applying 100 ms windowed correlations, the authors reveal a two-component activation–inhibition structure that accounts for most variance and shows robust cross-frequency coupling across subjects. The main contributions are (i) demonstrating distinct spatial patterns for $\Gamma$ activation and $\beta$ inhibition that are not captured by grid-averaged analyses, (ii) identifying a two-component built representation that closely tracks speech movement dynamics and can reconstruct activity with high fidelity, and (iii) linking these findings to broader motor control theories and potential non-invasive decoding approaches. This work advances understanding of cortical speech control, with practical implications for EEG-based speech decoding and neuromodulation, while acknowledging that ECoG's invasiveness limits subject diversity.

Abstract

I compute the average trial-by-trial power of band-limited speech activity across epochs of multi-channel high-density electrocorticography (ECoG) recorded from multiple subjects during a consonant-vowel speaking task. I show that previously seen anti-correlations of average beta frequency activity (12-35 Hz) to high-frequency gamma activity (70-140 Hz) during speech movement are observable between individual ECoG channels in the sensorimotor cortex (SMC). With this I fit a variance-based model using principal component analysis to the band-powers of individual channels of session-averaged ECoG data in the SMC and project SMC channels onto their lower-dimensional principal components. Spatiotemporal relationships between speech-related activity and principal components are identified by correlating the principal components of both frequency bands to individual ECoG channels over time using windowed correlation. Correlations of principal component areas to sensorimotor areas reveal a distinct two-component activation-inhibition-like representation for speech that resembles distinct local sensorimotor areas recently shown to have complex interplay in whole-body motor control, inhibition, and posture. Notably the third principal component shows insignificant correlations across all subjects, suggesting two components of ECoG are sufficient to represent SMC activity during speech movement.

Figures (18)

• Figure 1: SMC indices included by subject (precentral gyrus: light red; post-central gyrus: dark red; outside: grey) on a high-density 256-channel ECoG grid. Invalid ("bad") channels having noise and/or seizure-related (ictal) artifacts during session recordings (orange) are excluded from analysis. Reproduced from Easthope et al. EasthopeShamei2023, Supplementary Figure S1.
• Figure 2: Channel-by-channel bandpowers (256 channels) in a speech trial sample showing a heterogeneity of channel contribution to average ECoG channel bandpower.
• Figure 3: Average ECoG session $\Gamma$ bandpower summed by ECoG grid channel; "bright spots" (yellow) identify high-activation areas inside (coloured) and outside (coloured, reduced alpha) the SMC: $\Gamma$ activations in the SMC are localized robustly across subjects to the lower pre-central gyrus consistent with existing somatotopic models of tongue, jaw, and larynx activations as part of speech control.
• Figure 4: Average ECoG session $\Gamma$ bandpower by channel at "peak" or maximum (top row; maximum power) and "valley" or minimum (bottom row; minimum power) activation times; "bright spots" (yellow) identify high-activation areas inside (coloured) and outside (coloured, reduced alpha) in the SMC.
• Figure 5: Average ECoG session $\beta$ bandpower summed by ECoG grid channel; "bright spots" (yellow) identify high-activation areas inside (coloured) and outside (coloured, reduced alpha) the SMC: $\beta$ activations in the SMC appear less localized than $\Gamma$ activations without a clear activation "centre" away from the lower pre-central gyrus suggesting somatotopic differences between $\beta$ and $\Gamma$ activations.