Differential Effects of Brain Disorders on Structural and Functional Connectivity

Area of Science:

• Neuroimaging and computational neuroscience
• Graph theory applications in neuroscience
• Biostatistics and machine learning for biological data

Background:

• Neurological and psychiatric disorders often involve altered brain connectivity, but differentiating impacts on structural and functional connectivity is challenging.
• Existing methods struggle to quantitatively compare the effects of disease on diverse connectivity measures.
• Understanding these differences is crucial for diagnosing and treating brain conditions.

Purpose of the Study:

• To develop a novel statistical framework for quantifying and comparing the group-level effects of brain disease on structural and functional connectivity.
• To introduce a common representation space and a new test statistic for assessing differential impacts on connectivity types.
• To validate the proposed methods using a murine model of Agenesis of the Corpus Callosum (ACC).

Main Methods:

• Utilized a graph representation of brain networks to model structural and functional connectivity.
• Introduced a kernel two-sample test to assess group differences between patients and controls.
• Developed a novel test statistic within a common representation space to evaluate differential disease effects on connectivity types.

Main Results:

• The proposed methods successfully discriminated between BTBR mice (ACC model) and normo-callosal B6 mice, indicating ACC affects both connectivity types.
• The novel test statistic revealed that structural connectivity is significantly more affected by ACC than functional connectivity.
• Functional connectivity demonstrates a robust topology capable of withstanding substantial alterations in its structural substrate.

Conclusions:

• The developed statistical approach effectively quantifies and compares the impact of neurological disorders on distinct brain connectivity measures.
• Structural connectivity is more vulnerable to Agenesis of the Corpus Callosum than functional connectivity in the studied murine model.
• This work provides a valuable tool for understanding disease mechanisms and potentially informing therapeutic strategies in neurological conditions.