How and how not to correct for CSF-contamination in diffusion MRI

Metzler-Baddeley, Claudia, O'Sullivan, Michael J., Bells, Sonya, Pasternak, Ofer and Jones, Derek K. (2012) How and how not to correct for CSF-contamination in diffusion MRI. NeuroImage, 59 2: 1394-1403. doi:10.1016/j.neuroimage.2011.08.043


Author Metzler-Baddeley, Claudia
O'Sullivan, Michael J.
Bells, Sonya
Pasternak, Ofer
Jones, Derek K.
Title How and how not to correct for CSF-contamination in diffusion MRI
Journal name NeuroImage   Check publisher's open access policy
ISSN 1053-8119
1095-9572
Publication date 2012-01-16
Sub-type Article (original research)
DOI 10.1016/j.neuroimage.2011.08.043
Open Access Status Not yet assessed
Volume 59
Issue 2
Start page 1394
End page 1403
Total pages 10
Place of publication Amsterdam, Netherlands
Publisher Elsevier
Language eng
Subject 2808 Neurology
2805 Cognitive Neuroscience
Abstract Diffusion MRI is used extensively to investigate changes in white matter microstructure related to brain development and pathology. Ageing, however, is also associated with significant white and grey matter loss which in turn can lead to cerebrospinal fluid (CSF) based partial volume artefacts in diffusion MRI metrics. This is especially problematic in regions prone to CSF contamination, such as the fornix and the genu of corpus callosum, structures that pass through or close to the ventricles respectively. The aim of this study was to model the effects of CSF contamination on diffusion MRI metrics, and to evaluate different post-acquisition strategies to correct for CSF-contamination: Controlling for whole brain volume and correcting on a voxel-wise basis using the Free Water Elimination (FWE) approach. Using the fornix as an exemplar of a structure prone to CSF-contamination, corrections were applied to tract-specific and voxel-based [tract based spatial statistics (TBSS)] analyses of empirical DT-MRI data from 39 older adults (53-93. years of age). In addition to significant age-related decreases in whole brain volume and fornix tissue volume fraction, age was also associated with a reduction in mean fractional anisotropy and increase in diffusivity metrics in the fornix. The experimental data agreed with the simulations in that diffusivity metrics (mean diffusivity, axial and radial diffusivity) were more prone to partial volume CSF-contamination errors than fractional anisotropy. After FWE-based voxel-by-voxel partial volume corrections, the significant positive correlations between age and diffusivity metrics, in particular with axial diffusivity, disappeared whereas the correlation with anisotropy remained. In contrast, correcting for whole brain volume had little effect in removing these spurious correlations. Our study highlights the importance of correcting for CSF-contamination partial volume effects in the structures of interest on a voxel-by-voxel basis prior to drawing inferences about underlying changes in white matter structures and have implications for the interpretation of many recent diffusion MRI results in ageing and disease.
Formatted abstract
Diffusion MRI is used extensively to investigate changes in white matter microstructure related to brain development and pathology. Ageing, however, is also associated with significant white and grey matter loss which in turn can lead to cerebrospinal fluid (CSF) based partial volume artefacts in diffusion MRI metrics. This is especially problematic in regions prone to CSF contamination, such as the fornix and the genu of corpus callosum, structures that pass through or close to the ventricles respectively. The aim of this study was to model the effects of CSF contamination on diffusion MRI metrics, and to evaluate different post-acquisition strategies to correct for CSF-contamination: Controlling for whole brain volume and correcting on a voxel-wise basis using the Free Water Elimination (FWE) approach. Using the fornix as an exemplar of a structure prone to CSF-contamination, corrections were applied to tract-specific and voxel-based [tract based spatial statistics (TBSS)] analyses of empirical DT-MRI data from 39 older adults (53–93 years of age). In addition to significant age-related decreases in whole brain volume and fornix tissue volume fraction, age was also associated with a reduction in mean fractional anisotropy and increase in diffusivity metrics in the fornix. The experimental data agreed with the simulations in that diffusivity metrics (mean diffusivity, axial and radial diffusivity) were more prone to partial volume CSF-contamination errors than fractional anisotropy. After FWE-based voxel-by-voxel partial volume corrections, the significant positive correlations between age and diffusivity metrics, in particular with axial diffusivity, disappeared whereas the correlation with anisotropy remained. In contrast, correcting for whole brain volume had little effect in removing these spurious correlations. Our study highlights the importance of correcting for CSF-contamination partial volume effects in the structures of interest on a voxel-by-voxel basis prior to drawing inferences about underlying changes in white matter structures and have implications for the interpretation of many recent diffusion MRI results in ageing and disease.
Keyword Ageing
Atrophy
Diffusion MRI
Fornix
Partial volume effects
Q-Index Code C1
Q-Index Status Provisional Code
Institutional Status Non-UQ

Document type: Journal Article
Sub-type: Article (original research)
Collection: Queensland Brain Institute Publications
 
Versions
Version Filter Type
Citation counts: TR Web of Science Citation Count  Cited 111 times in Thomson Reuters Web of Science Article | Citations
Scopus Citation Count Cited 122 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Created: Thu, 08 Jun 2017, 13:59:00 EST by Kirstie Asmussen on behalf of Learning and Research Services (UQ Library)