Exploring intrinsic networks and their interactions using group wise temporal sparse coding

Recent resting state fMRI (rsfMRI) studies have shown that analysis of spontaneous activities may reveal intrinsic functional organization of the human brain. Increasing evidence has demonstrated that the human brain is organized as networks which dynamically interact with each other to realize brai...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings (International Symposium on Biomedical Imaging) pp. 74 - 77
Main Authors: Ge, Fangfei, Lv, Jinglei, Hu, Xintao, Guo, Lei, Han, Junwei, Zhao, Shijie, Liu, Tianming
Format: Conference Proceeding
Language:English
Published: IEEE 01.04.2018
Subjects:
Correlation
Dictionaries
dynamic interactions
Encoding
Functional magnetic resonance imaging
group-wise
intrinsic networks
Matrix decomposition
rsfMRI
Sparse matrices
temporal sparse coding
Training
ISSN:1945-8452
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recent resting state fMRI (rsfMRI) studies have shown that analysis of spontaneous activities may reveal intrinsic functional organization of the human brain. Increasing evidence has demonstrated that the human brain is organized as networks which dynamically interact with each other to realize brain functions. However, it is still challenging to model intrinsic networks and their dynamic interactions simultaneously. In this paper, we propose a novel group-wise temporal sparse coding (GTSC) method on rsfMRI data to address the challenge. Specifically, brain volume at each time point of rsfMRI is rearranged into a sample vector. After pooling all these sample vectors from multiple time points and multiple subjects as a training set, the dictionary learning and sparse coding method is employed to learn a set of spatial networks. Coded in the associated coefficient matrix, these networks are sparsely integrated at each time point while dynamically interacting along the time line. Experiment results have shown that our method is capable of detecting well-recognized intrinsic brain networks, and revealing their dynamic interactions simultaneously.
ISSN:1945-8452
DOI:10.1109/ISBI.2018.8363526