Conscious Perception as Integrated Information Patterns in Human Electrocorticography

Area of Science:

• Neuroscience
• Cognitive Science
• Information Theory

Background:

• Distinguishing neural processing linked to conscious perception from unconscious processing is a key challenge in neuroscience.
• Understanding the neural basis of consciousness is crucial for advancing cognitive science.

Purpose of the Study:

• To investigate if a hierarchical structure of causal interactions between neuronal populations correlates with conscious perception.
• To test the hypothesis that integrated information patterns, derived from Integrated Information Theory (IIT), reflect conscious awareness.

Main Methods:

• Utilized intracranial electrocorticography (ECoG) recordings from six human neurosurgical patients.
• Employed continuous flash suppression (CFS) and backward masking (BM) to manipulate conscious perception.
• Computed integrated information patterns and compared them with mutual information and entropy measures.

Main Results:

• Object-sensitive cortical areas showed a correspondence between conscious percepts and integrated information patterns.
• Unsupervised classification revealed that visual experiences clustered with integrated information, outperforming other information measures.
• This indicates that the structure of integrated information is specifically related to conscious object perception.

Conclusions:

• Locally integrated information plays a significant role in the neural substrate of conscious object perception.
• The findings support the potential of integrated information patterns as a marker for conscious awareness.
• This research offers new insights into the neural mechanisms underlying subjective experience.