Beyond Functional Connectivity: Investigating Networks of Multivariate Representations

Area of Science:

• Neuroscience
• Cognitive Science
• Medical Imaging

Background:

• Functional Connectivity (FC) studies have dominated brain interaction research for over 20 years.
• FC traditionally relies on univariate responses, potentially missing complex neural information.
• Thousands of FC studies investigate healthy brains and neurological disorders.

Purpose of the Study:

• To review the emergence of advanced multivariate and nonlinear methods for studying brain region interactions.
• To highlight the limitations of traditional univariate approaches in functional connectivity.
• To introduce novel techniques that capture richer information from brain activity patterns.

Main Methods:

• Discussion of recent advancements in multivariate functional Magnetic Resonance Imaging (fMRI) analyses.
• Exploration of nonlinear methods for assessing brain region interactions.
• Contrast between univariate and multivariate approaches to brain connectivity.

Main Results:

• Multivariate methods reveal information encoded in response patterns, often lost in univariate analyses.
• New techniques offer enhanced sensitivity to fluctuations in multivariate neural information.
• These methods enable a more nuanced understanding of how brain regions interact.

Conclusions:

• The field is shifting towards more sophisticated methods for analyzing brain connectivity.
• Multivariate and nonlinear approaches provide a deeper understanding of brain function and dysfunction.
• Future research can leverage these advanced techniques to explore the 'how' of brain interactions, not just the 'if'.