Modeling Consciousness with Integrated Information Theory
Keywords
Loading...
Authors
Issue Date
2024-07-01
Language
en
Document type
Journal Title
Journal ISSN
Volume Title
Publisher
Title
ISSN
Volume
Issue
Startpage
Endpage
DOI
Abstract
Understanding the neural basis of consciousness is one of the most profound challenges in both phi losophy and neuroscience. Integrated Information Theory (IIT) offers a promising framework for
exploring the mechanisms of consciousness. In this thesis, we implement IIT in high-dimensional
wide-field calcium imaging of mouse dorsal cortex across wake, sleep, and anesthesia states. Using
synchronized EEG and EMG signals, neural calcium activity was monitored and analyzed using ei ther anatomical regions or k-Means clusters as nodes for IIT calculations. The Φ-values, a measure of
integrated information, were calculated for different brain states and compared to test the hypothesis
that Φ-values are higher in conscious states (e.g. wake and recovery) than in unconscious states (e.g.
NREM sleep and anesthesia). The results demonstrate significant differences in Φ-values between
conscious and unconscious states, supporting the application of IIT in empirical neuroscience. How ever, limitations include the high computational demands and potential loss of detailed information
due to dimensionality reduction. To both retain the principles of IIT and a higher granularity, future
directions could utilize causal machine learning and attractor modeling for neural networks. Still,
this study bridges theoretical models and empirical data, contributing to a deeper understanding of
the neural substrates of consciousness and offering potential implications for neurology and cognitive
science.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen