An intracranial EEG study on auditory deviance detection.

Poster Session E - Monday, April 15, 2024, 2:30 – 4:30 pm EDT, Sheraton Hall ABC

Alejandro Omar Blenkmann¹ (ablenkmann@gmail.com), Vegard Volehaugen¹, Vinicius Rezende Carvahlo¹, Sabine Leske¹, Anaïs Llorens², Ingrid Funderud³, Santiago Collavini⁴, Pål Gunnar Larsson³, Jugoslval Ivanovic³, Tristan Bekinschtein⁵, Silvia Kochen⁴, Robert Knight⁶, Tor Endestad¹, Anne-Kristin Solbakk¹; ¹University of Oslo, ²FEMTO-ST institute, France, ³Oslo University Hospital, ⁴ENYS-CONICET, Argentina, ⁵University of Cambridge, ⁶University of California at Berkeley

The neural network underlying human auditory deviance detection is not fully understood. To address this, we recorded SEEG from 31 adult patients with drug-resistant epilepsy who had depth electrodes implanted in all brain lobes. The local Research Ethics Committee approved the study, and patients gave informed written consent to participate. Patients passively heard a stream of bilaterally presented tones while reading. We used the Optimum-1 paradigm [1], which consisted of 300 standard tones interleaved with 300 randomly presented deviant tones per block (3 to 10 blocks per patient). Deviant tones differed from standards in: 1) intensity (louder or softer), 2) frequency (higher or lower), 3) sound source location (right or left), 4) shorter duration, or 5) silent gap in the middle. Electrode coordinates were obtained from MRI and CT images using the iElectrodes toolbox. Non-epileptic channels were bipolar referenced (n=2041), and ERPs (0.1-30 Hz) and high-frequency band activity (HFA, 75-145 Hz) were extracted. Significant ERP responses to tones compared to the baseline period were observed primarily in temporal and insular areas, and also included the hippocampus, amygdala, and frontal, parietal, and cingulate cortices (1238 channels, FDR corrected). Deviance detection channels (409 channels, cluster-based permutation statistics) primarily showed in temporal, insular, pericentral, and prefrontal cortices, among other regions. HFA responses showed a similar profile but in a reduced proportion of channels. Our results show the spatiotemporal dynamics of a distributed brain network supporting auditory processing and deviance detection. [1] Näätänen et al., 2004, DOI:10.1016/j.clinph.2003.04.001.

Topic Area: PERCEPTION & ACTION: Audition