Supramodal and modality-specific neural information supports multi-feature prediction errors across cortical levels.

Area of Science:

• Neuroscience
• Cognitive Science
• Sensory Integration

Background:

• Predictive coding theory suggests the brain anticipates sensory input, generating prediction errors when reality mismatches expectations.
• Most research examines prediction errors from single sensory feature violations, not complex, multi-modal events typical of natural environments.

Purpose of the Study:

• To investigate how the brain processes prediction errors arising from violations across multiple sensory features and modalities.
• To explore the neural mechanisms and temporal dynamics of integrating complex sensory expectation violations.

Main Methods:

• A hierarchical oddball paradigm with auditory and somatosensory stimuli was used in 30 participants.
• High-density electroencephalography (EEG) recorded brain activity during single- and double-deviant stimulus presentation.
• Temporal decoding and effective connectivity analyses were performed to examine brain responses and interactions.

Main Results:

• Both single- and double-deviants elicited sustained supramodal activation patterns.
• Double-deviants uniquely triggered a supramodal response beginning at 100 ms post-stimulus.
• Effective connectivity revealed shared interhemispheric frontal interactions and distinct modality-specific pathways.

Conclusions:

• Multi-feature prediction errors engage rapid supramodal integration and hierarchically organized, modality-specific neural pathways.
• The brain flexibly integrates multiple sensory expectation violations at various cortical processing levels.
• Findings offer new insights into the neural architecture supporting predictive perception and multi-sensory integration.