Meeting Banner
Abstract #2098

Effects of Low Pass Filtering and Autocorrelation on Resting State FMRI as Investigated Using a Regression Model in SPM8

Kayako Matsuo1, Yung-Chin Hsu2, Wen-Yih I. Tseng1

1National Taiwan University College of Medicine, Taipei, Taiwan; 2Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Taipei, Taiwan


We examined an interaction between the low pass filtering (LPF) and the first degree autocorrelation (AR(1)) on SPM8 platform using resting state fMRI. We set 2 different procedures for the same data: with or without an LPF at 0.1Hz. Time courses were extracted from 3 cognitive mode network (CMN) regions as well as 3 default mode network (DMN) regions. We set 2 different GLMs for both of LPF and no LPF data: with or without AR(1). The LPF and the AR(1) acted in an opposite way with each other; specifically, the LPF enhanced DMN activity, whereas the AR(1) decreased.