Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Predictive Control and Self-Monitoring in Covert and Overt Speech Explored with Intracranial Directed Connectivity

Poster Session C - Sunday, March 8, 2026, 5:00 – 7:00 pm PDT, Fairview/Kitsilano Ballrooms

Taisha Donnelly¹ (taisha.donnelly@univ-grenoble-alpes.fr), Ioana Mîndruță^2,3, Andrei Barborica⁴, Andrei-Alexandru Vasiliu⁴, Romain Grandchamp¹, Jean-Baptiste Eichenlaub^1,5, Irina Oane^2,3, Monica Baciu¹, Hélène Lœvenbruck¹; ¹Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, LPNC, Grenoble, France, ²Epilepsy Monitoring Unit, Department of Neurology, Emergency University Hospital Bucharest, Bucharest, Romania, ³Department of Clinical Neuroscience, Carol Davila University of Medicine and Pharmacy Bucharest, Bucharest, Romania, ⁴Department of Physics, University of Bucharest, Bucharest, Romania, ⁵Institut Universitaire de France (IUF), Paris, France

Predictive control theories propose that speech motor planning generates auditory predictions via efference copies, which are internally monitored to support fluent production and, in inner speech, the phenomenology of the inner voice. High-gamma peaks (80–150Hz) during covert and overt picture naming suggest a consistent temporal sequence in which motor planning (~400ms) precedes auditory prediction-related activity (~500ms), followed by motor execution and feedback monitoring (>1000ms). To characterize directed information flow during internal monitoring of auditory predictions, we conducted spectral Granger causality (GC) analyses using a sliding-window approach. Alpha (8-12Hz), beta (13–30Hz), and low-gamma (30–70Hz) bands were evaluated for auditory–motor interactions between the premotor cortex (PMC), superior temporal gyrus (STG), and, given its role in salience detection and conscious attention, the anterior insula. Analyses focused on one patient with electrodes in all regions of interest who performed covert and overt picture naming during stereo-EEG recording as part of a presurgical evaluation for focal drug-resistant epilepsy. The insula appeared to function as a monitoring or integrative hub, as following peaks in local high-gamma activity (e.g., in PMC and STG), low-frequency directed influence flowed predominantly from the insula toward these regions. Reduced low-gamma GC originating from the PMC was observed for covert relative to overt speech during speech planning windows, which may reflect movement inhibition. Together, these findings support efference-copy–based auditory prediction during speech planning and suggest a role for the anterior insula in cascaded internal monitoring processes preceding speech production.

Topic Area: LANGUAGE: Other