Schizophrenia Shows Disrupted Links between Brain Volume and Dynamic Functional Connectivity

Area of Science:

• Neuroimaging
• Psychiatric Disorders
• Computational Neuroscience

Background:

• Schizophrenia is characterized by brain abnormalities that may be better understood using multimodal neuroimaging.
• Independent analyses of functional and structural brain data capture only partial information.

Purpose of the Study:

• To propose a novel framework for fusing functional magnetic resonance imaging (fMRI) and gray matter (GM) data in schizophrenia.
• To investigate the links between structural and time-varying functional connectivity changes in schizophrenia.

Main Methods:

• A multimodal fusion framework combining multiset canonical correlation analysis (mCCA) and joint ICA (jICA).
• Features included dynamic functional network connectivity (dFNC) maps and GM intensities from 151 schizophrenia patients and 163 healthy controls.
• dFNC was estimated using sliding window correlations, temporal ICA, and spatio-temporal regression.

Main Results:

• The fusion approach successfully linked changes in GM volumes and dFNC in schizophrenia patients.
• Significant group differences were found in GM regions including the superior parietal lobule, precuneus, and insula.
• Aberrations in inter-regional functional connectivity strength were also identified.

Conclusions:

• Multimodal data fusion reveals disrupted structure-function relationships in schizophrenia.
• This approach offers a more comprehensive understanding of schizophrenia than single-modality analyses.
• Future work will focus on replicating these findings in larger datasets.