Constructing High-Order Dynamic Functional Connectivity Networks from Resting-State fMRI for Brain Dementia Identification

Functional connectivity (FC) networks with the restingstate functional magnetic resonance imaging (rs-fMRI) help advance our understanding of brain disorders, such as Alzheimer’s disease (AD) and its prodromal stage, i.e., mild cognitive impairment (MCI). Recent studies have shown that FC networks demonstrate significant dynamic changes even in the resting state. However, previous studies typically focus on model the low-order (e.g., second-order) dynamics, without exploring the high-order dynamic properties of FC networks. In this paper, we propose to build a high-order dynamic functional connectivity network (hoDFCN) from the second-order FC networks, and define two novel measures to characterize the temporal and spatial variability of hoDFCN. Furthermore, we employ both spatial and temporal variability features for brain disease classification. Experimental results on 149 subjects with baseline resting-state functional MRI (rs-fMRI) data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) suggest the effectiveness of our proposed method in brain dementia identification.

[1]  N. Tzourio-Mazoyer,et al.  Automated Anatomical Labeling of Activations in SPM Using a Macroscopic Anatomical Parcellation of the MNI MRI Single-Subject Brain , 2002, NeuroImage.

[2]  Daniel L. Rubin,et al.  Network Analysis of Intrinsic Functional Brain Connectivity in Alzheimer's Disease , 2008, PLoS Comput. Biol..

[3]  Tom Vercauteren,et al.  Diffeomorphic demons: Efficient non-parametric image registration , 2009, NeuroImage.

[4]  Stefano Panzeri,et al.  The impact of high-order interactions on the rate of synchronous discharge and information transmission in somatosensory cortex , 2009, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[5]  Daoqiang Zhang,et al.  Multimodal classification of Alzheimer's disease and mild cognitive impairment , 2011, NeuroImage.

[6]  Martin A. Lindquist,et al.  Dynamic connectivity regression: Determining state-related changes in brain connectivity , 2012, NeuroImage.

[7]  David T. Jones,et al.  Non-Stationarity in the “Resting Brain’s” Modular Architecture , 2012, PloS one.

[8]  Hillary D. Schwarb,et al.  Short‐time windows of correlation between large‐scale functional brain networks predict vigilance intraindividually and interindividually , 2013, Human brain mapping.

[9]  David A. Leopold,et al.  Dynamic functional connectivity: Promise, issues, and interpretations , 2013, NeuroImage.

[10]  David J. Sharp,et al.  Network dysfunction after traumatic brain injury , 2014, Nature Reviews Neurology.

[11]  Daoqiang Zhang,et al.  Manifold regularized multitask feature learning for multimodality disease classification , 2015, Human brain mapping.

[12]  Martin A. Lindquist,et al.  Assessing uncertainty in dynamic functional connectivity , 2017, NeuroImage.

[13]  Dinggang Shen,et al.  Integration of temporal and spatial properties of dynamic connectivity networks for automatic diagnosis of brain disease , 2018, Medical Image Anal..

[14]  Daoqiang Zhang,et al.  Spatial-Temporal Dependency Modeling and Network Hub Detection for Functional MRI Analysis via Convolutional-Recurrent Network , 2019, IEEE Transactions on Biomedical Engineering.