Latent source mining in FMRI via restricted Boltzmann machine

Blind source separation (BSS) is commonly used in functional magnetic resonance imaging (fMRI) data analysis. Recently, BSS models based on restricted Boltzmann machine (RBM), one of the building blocks of deep learning models, have been shown to improve brain network identification compared to conv...

Full description

Saved in:
Bibliographic Details
Published in:Human brain mapping Vol. 39; no. 6; pp. 2368 - 2380
Main Authors: Hu, Xintao, Huang, Heng, Peng, Bo, Han, Junwei, Liu, Nian, Lv, Jinglei, Guo, Lei, Guo, Christine, Liu, Tianming
Format: Journal Article
Language:English
Published: United States John Wiley & Sons, Inc 01.06.2018
John Wiley and Sons Inc
Subjects:
Algorithms
blind source separation
Brain
Brain - diagnostic imaging
Brain Mapping
Complexity
Data analysis
Data mining
Data processing
Datasets as Topic - statistics & numerical data
Functional magnetic resonance imaging
Humans
Image Processing, Computer-Assisted
Independent component analysis
Machine learning
Magnetic Resonance Imaging
Models, Neurological
Neuroimaging
Oxygen - blood
restricted Boltzmann machine
Signal processing
Task Performance and Analysis
Training
blind source separation
functional magnetic resonance imaging
restricted Boltzmann machine
ISSN:1065-9471, 1097-0193, 1097-0193
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Blind source separation (BSS) is commonly used in functional magnetic resonance imaging (fMRI) data analysis. Recently, BSS models based on restricted Boltzmann machine (RBM), one of the building blocks of deep learning models, have been shown to improve brain network identification compared to conventional single matrix factorization models such as independent component analysis (ICA). These models, however, trained RBM on fMRI volumes, and are hence challenged by model complexity and limited training set. In this article, we propose to apply RBM to fMRI time courses instead of volumes for BSS. The proposed method not only interprets fMRI time courses explicitly to take advantages of deep learning models in latent feature learning but also substantially reduces model complexity and increases the scale of training set to improve training efficiency. Our experimental results based on Human Connectome Project (HCP) datasets demonstrated the superiority of the proposed method over ICA and the one that applied RBM to fMRI volumes in identifying task‐related components, resulted in more accurate and specific representations of task‐related activations. Moreover, our method separated out components representing intermixed effects between task events, which could reflect inherent interactions among functionally connected brain regions. Our study demonstrates the value of RBM in mining complex structures embedded in large‐scale fMRI data and its potential as a building block for deeper models in fMRI data analysis.
Bibliography:Funding information
National Key R&D Program of China, Grant/Award Number: 2017YFB1002201 National Natural Science Foundation of China (NSFC), Grant/Award Numbers: 61473234, 61333017, 61522207; Fundamental Research Funds for the NIH Career Award, Grant/Award Number: EB006878, NSF Career Award, Grant/Award Numbers: IIS‐1149260, NIH R01 DA033393, NIH R01 AG042599, NSF CBET‐1302089, NSF BCS‐1439051, NSF DBI‐1564736
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
Funding information National Key R&D Program of China, Grant/Award Number: 2017YFB1002201 National Natural Science Foundation of China (NSFC), Grant/Award Numbers: 61473234, 61333017, 61522207; Fundamental Research Funds for the NIH Career Award, Grant/Award Number: EB006878, NSF Career Award, Grant/Award Numbers: IIS‐1149260, NIH R01 DA033393, NIH R01 AG042599, NSF CBET‐1302089, NSF BCS‐1439051, NSF DBI‐1564736
ISSN:1065-9471
1097-0193
1097-0193
DOI:10.1002/hbm.24005