Neural complexity and structural connectivity

Area of Science:

• Computational neuroscience
• Information theory
• Network science

Background:

• Neural complexity, proposed by Tononi, measures information integration in neural networks using mutual information.
• The original measure faces computational challenges due to exponential scaling with system size.

Purpose of the Study:

• To develop an efficient approximation of Tononi's neural complexity measure for Gaussian models.
• To elucidate the relationship between neural system complexity and structural connectivity.

Main Methods:

• Developed an approximation for neural complexity within a popular Gaussian model framework.
• Applied the approximation to continuous-time neural processes.
• Analyzed scaling properties and computational cost compared to the original measure.

Main Results:

• The approximation accurately captures neural complexity for weakly coupled systems.
• The method demonstrates polynomial scaling with system size, offering significant computational advantages.
• Established a clear link between neural complexity and structural connectivity in Gaussian models.

Conclusions:

• The proposed approximation provides a computationally feasible way to assess neural complexity.
• This work enhances understanding of how network structure influences information processing in neural systems.