Comparison and Dynamic interaction between Auditory Cortex and Prefrontal Cortex of Behaving Monkeys during Novelty Detection

The ability to detect deviations from expected sensory input is critical for adaptive behavior. Here, we investigated the neural dynamics of novelty detection in the auditory cortex (AC) and prefrontal cortex (PFC) of macaques performing an auditory oddball task. Using high-density electrocorticography and Granger causality analyses (GCA), we identified distinct repetition-related signals in both regions: the AC showed both suppression and facilitation to repeated stimuli, while the PFC exhibited robust enhancement, particularly in low-frequency (2 Hz) oscillatory activity. These predictive responses were accompanied by bidirectional AC-PFC coupling in the delta band, reflecting a temporally coordinated predictive state. Deviant tones elicited early responses in the AC, followed by PFC activation, and were associated with increased feedforward and feedback connectivity across delta, theta, and gamma bands. Task engagement amplified both prediction and prediction-error signals, enhancing interareal interactions and behavioral selectivity. Our results reveal that the AC encodes early sensory regularities and violations, while the PFC supports integrative and state-dependent prediction, jointly forming a distributed cortical network for hierarchical predictive coding.