Copula linked parallel ICA jointly estimates linked structural and functional MRI brain networks.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biostatistics

Background:

• Brain imaging techniques like functional magnetic resonance imaging (fMRI) and structural MRI (sMRI) offer complementary insights into brain structure and function.
• Current fusion methods often rely on pre-extracted fMRI features, neglecting the rich temporal dynamics within the 4D fMRI data.
• Observed similarities between resting-state fMRI networks and structural MRI networks provide a basis for integrated analysis.

Purpose of the Study:

• To introduce a novel method, copula linked parallel Independent Component Analysis (CLiP-ICA), for simultaneous fusion of fMRI and sMRI data.
• To leverage the full temporal information of fMRI data in the fusion process.
• To investigate the linkage between functional and structural brain networks using this new approach.

Main Methods:

• CLiP-ICA simultaneously estimates independent components and unmixing matrices for fMRI and sMRI.
• A copula model is employed to link the spatial sources derived from each imaging modality.
• The method was validated using simulated data and real-world fMRI and sMRI data from an Alzheimer's disease study.

Main Results:

• CLiP-ICA successfully identified significant linkages between fMRI and sMRI data.
• Linkages were observed in key brain regions and networks, including the cerebellum, sensorimotor areas, and the default mode network.
• The method demonstrated its ability to preserve temporal information from fMRI during the fusion process.

Conclusions:

• CLiP-ICA offers a powerful new approach for integrating multimodal brain imaging data (fMRI and sMRI).
• This method enhances our understanding of brain structure-function relationships by preserving temporal dynamics.
• The findings suggest potential applications in neurodegenerative disease research, such as Alzheimer's disease.