Back to Search
Start Over
A group-level comparison of volumetric and combined volumetric-surface normalization for whole brain analyses of myelin and iron maps.
- Source :
-
Magnetic resonance imaging [Magn Reson Imaging] 2018 Dec; Vol. 54, pp. 225-240. Date of Electronic Publication: 2018 Sep 01. - Publication Year :
- 2018
-
Abstract
- Quantitative MRI (qMRI) provides surrogate brain maps of myelin and iron content. After spatial normalization to a common standard brain space, these may be used to detect altered myelination and iron accumulation in clinical populations. Here, volumetric and combined volumetric and surface-based (CVS) normalization were compared to identify which procedure would afford the greatest sensitivity to inter-regional differences (contrast), and the lowest inter-subject variability (under normal conditions), of myelin- and iron-related qMRI parameters, in whole-brain group-level studies. Ten healthy volunteers were scanned twice at 3 Tesla. Three-dimensional T1-weighted, T2-weighted and multi-parametric mapping sequences for brain qMRI were used to map myelin and iron content over the whole brain. Parameter maps were spatially normalized using volumetric (DARTEL) and CVS procedures. Tissue probability weighting and isotropic Gaussian smoothing were integrated in DARTEL for voxel-based quantification (VBQ). Contrasts, coefficients of variations and sensitivity to detecting differences in the parameters were estimated in standard space for each approach on region of interest (ROI) and voxel-by-voxel bases. The contrast between cortical and subcortical ROIs with respectively different myelin and iron content was higher following CVS, compared to DARTEL-VBQ, normalization. Across cortical voxels, the inter-individual variability of myelin and iron qMRI maps were comparable between CVS (with no smoothing) and DARTEL-VBQ (with smoothing). CVS normalization of qMRI maps preserves higher myelin and iron contrast than DARTEL-VBQ over the entire brain, while exhibiting comparable variability in the cerebral cortex without extra smoothing. Thus, CVS may prove useful for detecting small microstructural differences in whole-brain group-level qMRI studies.<br /> (Copyright © 2018 Elsevier Inc. All rights reserved.)
- Subjects :
- Adult
Brain Stem
Cerebral Cortex
Healthy Volunteers
Humans
Image Processing, Computer-Assisted
Normal Distribution
Probability
Reproducibility of Results
Surface Properties
Young Adult
Brain diagnostic imaging
Brain Mapping methods
Iron chemistry
Magnetic Resonance Imaging
Myelin Sheath chemistry
Subjects
Details
- Language :
- English
- ISSN :
- 1873-5894
- Volume :
- 54
- Database :
- MEDLINE
- Journal :
- Magnetic resonance imaging
- Publication Type :
- Academic Journal
- Accession number :
- 30176374
- Full Text :
- https://doi.org/10.1016/j.mri.2018.08.021