A Bayesian approach for dynamic brain network Público

Abstract

Functional magnetic resonance imaging (fMRI) has been widely used in brain network research. Functional connectivity (FC) measures how different brain regions contact with each other. Recently there has been an increased interest in understanding the dynamic manner in the functional connectivity. Although sliding window method is still the most widely used one, because of its limitations in window size pick and interpretation, many researchers are trying to create new method. Dynamic Connectivity Regression (DCR) is a data-driven method to detect temporal change points in different brain regions. However, DCR may fail to detect some change points and is hard to detect rapid change in functional connectivity. In this paper, we introduce our Bayesian approach which combines both change point detection and Bayesian method to detect the number and positions of change points in FC simultaneously. Our method is based on the change point in precision matrix instead of the mean value of time series. Screening method like screening and ranking algorithm (SaRa) is also included in our method to increase the computation speed. Different choices of change points combinations are also provided to get an accurate estimation. Two simulation show that our method can provide a good estimation of positions when the number of change points is given. In addition, we provide an experiment data which can be used to validate our method.

Table of Contents

1. Introduction. 1

2. Methods. 3

2.1 Problem setup. 4

2.2 Change points method. 5

2.3 Screening method. 6

2.4 Dynamic Change point method. 6

2.5 Dynamic Connectivity Regression (DCR) method. 9

3. Simulation. 10

3.1 simulation design. 10

3.2 Experimental data. 11

4. Results. 12

5. Discussion. 16

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Biostatistics and Bioinformatics
Subfield / Discipline	Biostatistics - MPH & MSPH
Degree	MSPH
Submission	Master's Thesis
Language	English
Research Field	Biology, Biostatistics Biology, Neuroscience
Palavra-chave	Dynamic Functional Connectivity fMRI Change Point
Committee Chair / Thesis Advisor	Kundu, Suprateek , Emory University
Committee Members	Guo, Ying, Emory University

Última modificação

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	A Bayesian approach for dynamic brain network ()	2018-08-28 11:47:55 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions