Sources of signal fluctuations in functional magnetic resonance imaging at 7 Tesla

High-field magnetic resonance imaging (MRI) systems allow for critical improvements in image signal-to-noise ratio (SNR), potentially leading to higher sensitivity and spatial resolution for functional MRI (fMRI). 3D segmented echo volumar imaging (EVI) has recently been proposed for high resolution fMRI at ultra-high fields. It provides higher image SNR than standard 2D echo planar imaging (EPI), but is also thought to be more sensitive to physiological noise. This becomes especially important at higher field strengths, where increased signal fluctuations from correlated noise sources, namely physiological processes, have been observed. This work is focused on the analysis of fMRI data acquired at 7T using standard EPI and segmented EVI. A principal component analysis-based approach and a physiological regressor-based approach were investigated for correlated noise characterization and correction. Although increases in physiological noise from EPI to segmented EVI were found, both corrective approaches produced significant improvements in temporal SNR, explained signal information, and activation sensitivity for either acquisition technique, and especially for segmented EVI.