Share Email Print
cover

Proceedings Paper

Detecting brain dynamics during resting state: a tensor based evolutionary clustering approach
Author(s): Esraa Al-sharoa; Mahmood Al-khassaweneh; Selin Aviyente
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Human brain is a complex network with connections across different regions. Understanding the functional connectivity (FC) of the brain is important both during resting state and task; as disruptions in connectivity patterns are indicators of different psychopathological and neurological diseases. In this work, we study the resting state functional connectivity networks (FCNs) of the brain from fMRI BOLD signals. Recent studies have shown that FCNs are dynamic even during resting state and understanding the temporal dynamics of FCNs is important for differentiating between different conditions. Therefore, it is important to develop algorithms to track the dynamic formation and dissociation of FCNs of the brain during resting state. In this paper, we propose a two step tensor based community detection algorithm to identify and track the brain network community structure across time. First, we introduce an information-theoretic function to reduce the dynamic FCN and identify the time points that are similar topologically to combine them into a tensor. These time points will be used to identify the different FC states. Second, a tensor based spectral clustering approach is developed to identify the community structure of the constructed tensors. The proposed algorithm applies Tucker decomposition to the constructed tensors and extract the orthogonal factor matrices along the connectivity mode to determine the common subspace within each FC state. The detected community structure is summarized and described as FC states. The results illustrate the dynamic structure of resting state networks (RSNs), including the default mode network, somatomotor network, subcortical network and visual network.

Paper Details

Date Published: 24 August 2017
PDF: 14 pages
Proc. SPIE 10394, Wavelets and Sparsity XVII, 103940G (24 August 2017); doi: 10.1117/12.2274058
Show Author Affiliations
Esraa Al-sharoa, Michigan State Univ. (United States)
Mahmood Al-khassaweneh, Michigan State Univ. (United States)
Yarmouk Univ. (Jordan)
Selin Aviyente, Michigan State Univ. (United States)


Published in SPIE Proceedings Vol. 10394:
Wavelets and Sparsity XVII
Yue M. Lu; Dimitri Van De Ville; Manos Papadakis, Editor(s)

© SPIE. Terms of Use
Back to Top