Estimating Dynamic Connectivity States in fMRI Using Regime-Switching Factor Models

• Published in: IEEE transactions on medical imaging Vol. 37; no. 4; pp. 1011 - 1023
• Main Authors: Chee-Ming Ting, Ombao, Hernando, Samdin, S. Balqis, Salleh, Sh-Hussain
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.04.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Autoregressive processes; Brain; Brain - diagnostic imaging; Brain - physiology; Brain mapping; Brain modeling; Cluster Analysis; Clustering; Computer simulation; Connectome - methods; Covariance matrices; dynamic brain connectivity; Economic models; Estimation; factor analysis; fMRI; Functional magnetic resonance imaging; Hidden Markov models; Humans; Image Processing, Computer-Assisted - methods; large VAR models; Load modeling; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Markov chains; Models, Statistical; Neural networks; Neural Pathways - diagnostic imaging; Neuroimaging; Principal components analysis; Reactive power; Regime-switching models; Switching theory; Vector quantization; Windows (intervals)
• ISSN: 0278-0062; 1558-254X; 1558-254X
• DOI: 10.1109/TMI.2017.2780185

We consider the challenges in estimating the state-related changes in brain connectivity networks with a large number of nodes. Existing studies use the sliding-window analysis or time-varying coefficient models, which are unable to capture both smooth and abrupt changes simultaneously, and rely on ad-hoc approaches to the high-dimensional estimation. To overcome these limitations, we propose a Markov-switching dynamic factor model, which allows the dynamic connectivity states in functional magnetic resonance imaging (fMRI) data to be driven by lower-dimensional latent factors. We specify a regime-switching vector autoregressive (SVAR) factor process to quantity the time-varying directed connectivity. The model enables a reliable, data-adaptive estimation of change-points of connectivity regimes and the massive dependencies associated with each regime. We develop a three-step estimation procedure: 1) extracting the factors using principal component analysis, 2) identifying connectivity regimes in a low-dimensional subspace based on the factor-based SVAR model, and 3) constructing high-dimensional state connectivity metrics based on the subspace estimates. Simulation results show that our estimator outperforms K-means clustering of time-windowed coefficients, providing more accurate estimate of time-evolving connectivity. It achieves percentage of reduction in mean squared error by 60% when the network dimension is comparable to the sample size. When applied to the resting-state fMRI data, our method successfully identifies modular organization in the resting-state networks in consistency with other studies. It further reveals changes in brain states with variations across subjects and distinct large-scale directed connectivity patterns across states.