Schedule of Events | Symposia

Evaluating the Neurophysiological Evidence for Cross Modal Expectations using Alpha Oscillations

Poster Session F - Tuesday, April 1, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom
Also presenting in Data Blitz Session 3 - Saturday, March 29, 2025, 10:30 am – 12:00 pm EDT, Constitution A.

Soukhin Das¹ (skndas@ucdavis.edu), Mingzhou Ding², George (Ron) Mangun^1,3; ¹Department of Psychology and Center for Mind and Brain, University of California Davis, ²J Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, ³Department of Neurology, University of California Davis

Predictive coding is a general model of sensory processing in the brain that forms expectations about the likely causes of sensory input to predict future outcomes. Discrepancies between expectations and sensory inputs are propagated hierarchically as prediction errors to update these expectations. Recent studies suggest that alpha oscillations (8–12 Hz) are a spectral signature of predictive coding in different modalities. However, their distinct patterns during cross-modal sensory processing remain underexplored. Using a 2×2 cross-modal design, auditory (HEAR or SEE) and visual cues (H or S) indicated the target modality (visual/auditory). In 80% of trials, targets were presented in the expected modality, while 20% of trials involved targets in the unexpected modality. Participants distinguished the frequency of visual gratings or the tone of auditory stimuli irrespective of cue validity. Our findings revealed dissociative alpha desynchronization in central and occipital channels, guiding modality-specific expectations. Topographical analysis demonstrated modality-specific cortical activation, with stronger alpha suppression in occipital regions during correct responses compared to incorrect ones. Post-stimulus analysis revealed distinct alpha dynamics for expected versus unexpected targets. Expected targets exhibited sustained occipital alpha desynchronization, while unexpected targets showed transient alpha modulation, reflecting the neural cost of processing errors. Temporal generalization of alpha-power suggested sequential propagation of prediction error signals across cortical regions, highlighting efficient hierarchical updates to internal models. In summary, our study underscores the critical role of alpha oscillations in predictive coding, revealing distinct sensory and cognitive mechanisms underlying expectation and prediction error processing during cross-modal tasks.

Topic Area: PERCEPTION & ACTION: Multisensory