A temporal deconvolution algorithm for multiecho functional MRI

Project
Author
C. Gaudes
P. Bandettini
J. Gonzalez-Castillo
Keywords
biomedical MRI
Blood flow
BOLD fMRI signal
BOLD responses
BOLD signal
Brain
deconvolution
Functional magnetic resonance imaging
Generalized LASSO
inverse problem
inverse problems
ME-PFM
medical signal processing
Multiecho functional MRI
multiecho functional MRI data
multiecho paradigm free mapping
net magnetization
paradigm free mapping
previous deconvolution approaches
temporal deconvolution algorithm
Timing
transverse relaxation
Abstract
<p>This work introduces a novel method for the temporal deconvolution of the BOLD signal for multiecho functional MRI (fMRI) data: Multiecho Paradigm Free Mapping (ME-PFM). By solving an inverse problem with the Generalized LASSO, this method obtains voxelwise time-varying estimates of the changes in the transverse relaxation (R) and the net magnetization (So) of the fMRI signal without prior knowledge of the timings of the BOLD responses. Our results with experimental data demonstrate that the maps of R changes obtained with ME-PFM at the times of the stimulus trials closely resemble the maps obtained with a standard approach that is aware of the onsets and durations of the events. Compared with previous deconvolution approaches for BOLD fMRI, this method yields estimates of R changes in physiologically interpretable units, which is particularly relevant for the quantification of the dynamics of the BOLD fMRI signal during tasks and resting state.</p>
Year of Publication
2018
Date Published
April
DOI
10.1109/ISBI.2018.8363649