Title

Dynaconn: A Software for Dynamic Functional Connectivity Analysis of FMRI Data

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Science and Info Sys

Date of Award

Summer 2014

Abstract

Functional magnetic resonance imaging (fMRI) techniques have given researchers an amazing tool to elucidate brain dynamics in human subjects. Identifying areas in an fMRI course that have a high blood oxygenation level dependent (BOLD) signal can give insight into which brain regions are active during an fMRI scan. Repeated fMRI scans of the brain can give a temporal record of what brain regions are doing in terms of activity. A technique called functional connectivity (FC) can be used to identify simultaneous temporal correlation between different regions of brain activity. These functional connections of neural activity in the brain can be quite dynamic and should be investigated with dynamic tools. Dynamic functional connectivity (DFC) methods are used to measure how the temporal functional connections of neural activity change over short (10 to 50 sec) periods of time. Dynamic functional network connectivity (DFNC) is a specific type of DFC analysis where DFC is applied to brain networks such as those obtained by spatial independent component analysis (ICA).In this work, we developed a software called DynaConn, which is an easy-to-use Graphical User Interface (GUI) supported toolbox to perform DFC and DFNC analyses in MATLAB. DynaConn can read 4D fMRI data, different brain atlas files, which labels different anatomical regions and it can do co-registration between atlas and data. DFC analysis can be applied to any fMRI data directly or to networks found by ICA. ICA results can be evaluated to show the most significant connections among brain networks in individual or a group of subjects. DynaConn allows neuroscientists to identify functionally connected brain regions both visually and quantitatively. In the software, DFC among different brain regions or networks is calculated by computing the correlation among fMRI time courses using the sliding time window method. The window step size and width are adjustable through the visual interfaces. The dynamics of the connectivity can be plotted, along with any experimental tasks (stimuli). This process visualizes how DFC is modulated by task, if any.

Advisor

Mutlu Mete

Subject Categories

Computer Sciences | Physical Sciences and Mathematics

Link to Full Text
COinS