We consider challenges in modeling high-dimensional connectivity in brain networks whose number of nodes is large and are arranged with hierarchical and modular structure. We propose a multi-scale factor analysis (MSFA) model which partitions the massive data into a finite set of regional clusters. To achieve further dimension reduction, signals in each cluster are represented by a small number of latent factors. This enables a reliable and computationally efficient multi-scale analysis of both regional and global network connectivity. Simulation results show that the proposed MSFA estimator improves accuracy of connectivity estimation in high-dimensional settings. When applied to resting-state functional magnetic resonance \imaging (fMRI) data, our method identifies modular structure of resting-state networks at multiple scales, revealing interesting connectivity patterns between voxels, regions of interest (ROIs), and functional networks. This model will be illustrated to fMRI data collected while participants were watching a video.
This is joint work with Chee-Ming Ting (Univ Teknologi Malaysia and KAUST Saudi Arabia).