Complexity-based Encoded Information Quantification in Neurophysiological Recordings.

Area of Science:

• Neuroscience
• Information Theory
• Computational Neuroscience

Background:

• Brain activity exhibits distinct patterns during sleep, cognitive tasks, and action.
• Information theory provides a framework for analytically quantifying these brain states from neurophysiological recordings.
• Current methods using algorithmic information theory can discriminate brain states but struggle to quantify response similarity.

Purpose of the Study:

• To propose a novel method based on algorithmic information theory for directly quantifying the similarity between brain responses.
• To estimate encoded information through a compression-based scheme for direct statements on response similarity.
• To validate the proposed method against synthetic and real neurophysiological data.

Main Methods:

• Utilizing algorithmic information theory and a compression-based scheme to estimate encoded information.
• Applying the method to both synthetic and real neurophysiological data.
• Comparing the efficiency of the proposed method with the mutual information measure.

Main Results:

• The proposed method allows for direct quantification of similarity between neuronal responses.
• The compression-based scheme effectively estimates encoded information in brain activity.
• The method demonstrates precision in quantifying neuronal response similarity, especially for contrasting event types.

Conclusions:

• The developed compression-based method offers a significant advancement in analyzing brain activity by directly measuring response similarity.
• This approach enhances the capability of information theory in neuroscience for understanding brain mechanisms.
• The method is particularly valuable for experimental paradigms involving the comparison of different event-related brain responses.