The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Effect of scan time on resting state parameters
Author:	Patel, Dhruti
View Online:	njit-etd2013-043 (xii, 48 pages ~ 1.3 MB pdf)
Department:	Department of Biomedical Engineering
Degree:	Master of Science
Program:	Biomedical Engineering
Document Type:	Thesis
Advisory Committee:	Biswal, Bharat (Committee chair) Alvarez, Tara L. (Committee member) Di, Xin (Committee member)
Date:	2013-05
Keywords:	Local intrinsic brain activity Amplitude of low frequency fluctuations Fractional amplitude of low frequency fluctuations Acquisition time
Availability:	Unrestricted
Abstract:	In the past decade the interest in studying the spontaneous low-frequency fluctuations (LFF) in a resting-state brain has steadily grown. By measuring LFF (< 0.08 Hz) in blood-oxygen-level-dependent (BOLD) signals, resting-state functional magnetic resonance imaging (rs-fMRI) has proven to be a powerful tool in exploring brain network connectivity and functionality. Rs-fMRI data can be used to organize the brain into resting state networks (RSNs). In this thesis, rs-fMRI data are used to determine the minimum data acquisition time necessary to detect local intrinsic brain activity as a function of both the amplitude of low frequency fluctuations (ALFF) and the fractional amplitude of low frequency fluctuations (fALFF) in BOLD signals in healthy subjects. The data are obtained from 22 healthy subjects to use as a baseline for future rs-fMRI analysis. Voxel-wise analysis is performed on the whole brain, gray matter volume, and two previously established RSNs: the default mode network (DMN) and the visual system network, for all the subjects in this study. Pearson’s correlation coefficients (r-values) are calculated from each subject. The entire time series for one subject is divided into 31 subsections and the r-values are calculated between each consecutive subsection in a subject. In total, there are 30 r- values. To better understand what the results mean across subjects and within subjects Fisher transformations are applied to the 30 calculated r-values for each subject to get a normal z-distribution. The mean across 22 subjects’ z-values is calculated for group analysis. In the end, there are 30 mean values. Finally, an exponential curve fit model is calculated across the 22 subjects using the calculated mean values, and an asymptotic growth model is used to detect the minimum data acquisition time required to obtain both ALFF and fALFF of the BOLD signals at rest. The results show that the minimum time required to detect an ALFF and fALFF of the BOLD signals at rest is 12 and 13.33 minutes respectively. Future studies can focus on determining the minimum scanner time using similar analysis for different physiological states of the brain.