Structure supports function: Informing directed and dynamic functional connectivity with anatomical priors

Area of Science:

• Neuroscience
• Computational Neuroscience
• Brain Imaging

Background:

• Brain's functional networks dynamically change but are constrained by underlying structural connections.
• Integrating structural connectivity (SC) and functional connectivity (FC) using multimodal approaches remains a challenge.
• Existing methods have limitations in accurately estimating dynamic and directed FC.

Purpose of the Study:

• To develop and validate a new adaptive filter for estimating dynamic and directed functional connectivity (FC).
• To leverage structural connectivity (SC) information as priors to improve FC estimation.
• To enhance multimodal integration of brain structure and function data.

Main Methods:

• Proposed a novel adaptive filter incorporating structural connectivity (SC) information as priors.
• Tested the filter using rat epicranial recordings and human event-related EEG data.
• Utilized SC priors derived from tracer studies and diffusion tensor imaging metrics.

Main Results:

• The SC-informed filter refined estimates of directed FC, especially in low signal-to-noise ratio conditions.
• The filter promoted sparse functional networks that integrated information from both structure and function.
• Demonstrated intrinsic protection against SC-related false negatives and robustness against false positives.

Conclusions:

• The developed adaptive filter is a valuable tool for multimodal neuroimaging analysis.
• It improves the estimation of dynamic and directed functional connectivity by integrating structural priors.
• Offers enhanced accuracy and robustness for brain network analysis.