A comparison between voxel-based cortical thickness and voxel-based morphometry in normal aging.
Hutton C, Draganski B, Ashburner J, Weiskopf N. · 2009
Read A comparison between voxel-based cortical thickness and voxel-based morphometry in normal aging. by Hutton C, Draganski B, Ashburner J, Weiskopf N. (2009) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc2741580.txt.
Introduction
Hutton, Draganski, Ashburner, and Weiskopf compared in 2009 two MRI methods for measuring grey matter change in healthy aging: voxel-based cortical thickness and voxel-based morphometry. Both use standard brain scans, but one isolates cortical thickness while the other mixes thickness with surface area and folding. Regions with known age change were their test bed.
Overall the two methods agreed closely on where grey matter declines with age. The thickness method proved the more sensitive of the two, picking up age-linked decline more strongly. Since morphometry blends several anatomical features, using both together can separate true thinning from changes in area or folding. The pairing thus gives a fuller picture of how the aging cortex changes.
Catalog record
| CREATOR | Hutton C, Draganski B, Ashburner J, Weiskopf N. |
|---|---|
| YEAR | 2009 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A comparison between voxel-based cortical thickness and voxel-based morphometry in normal aging By Chloe Hutton, Bogdan Draganski, John Ashburner, Nikolaus Weiskopf Published in: Neuroimage DOI: 10.1016/j.neuroimage.2009.06.043 ABSTRACT Abstract The morphology of cortical grey matter is commonly assessed using T1-weighted MRI together with automated computerised methods such as voxel-based morphometry (VBM) and cortical thickness measures. In the presented study we investigate how grey matter changes identified using voxel-based cortical thickness (VBCT) measures compare with local grey matter volume changes identified using VBM. We use data from a healthy aging population to perform the comparison, focusing on brain regions where age-related changes have been observed in previous studies. Our results show that overall, in healthy aging, VBCT and VBM yield very consistent results but VBCT provides a more sensitive measure of age-associated decline in grey matter compared with VBM. Our findings suggest that while VBCT selectively investigates cortical thickness, VBM provides a mixed measure of grey matter including cortical surface area or cortical folding, as well as cortical thickness. We therefore propose that used together, these techniques can separate the underlying grey matter changes, highlighting the utility of combining these complementary methods. Keywords: Morphometry, VBM, Cortical thickness, Aging INTRODUCTION The morphometric analysis of magnetic resonance images (MRI) of the brain has become a widely used approach to investigate neuroanatomical correlates of both normal brain development and neurological disorders. Most typically, changes in grey matter have been assessed using T1-weighted images while changes in white matter are more typically explored using diffusion tensor imaging (DTI) or diffusion weighted imaging (DWI) e.g. Smith et al. (2006). Some of the most commonly used methods for investigating cortical grey matter in T1-weighted images include volumetric comparisons of manually, semi-automatically or automatically delineated neuroanatomical regions of interest, whole-brain voxel-based comparisons of grey matter and cortical surface-based comparisons of cortical thickness. A commonly used method for performing voxel-based comparisons of grey matter is known as voxel-based morphometry or VBM (Ashburner and Friston, 2000; Wright et al., 1995). This method generally relies on the MR images first being matched up in a common space, then, after correcting for intensity non-uniformities, voxels are classified as grey matter, white matter and cerebral spinal fluid (CSF). The value at each voxel in the resulting tissue segments can be thought of as representing the proportion of the corresponding tissue in that voxel. After smoothing the grey matter segments to create normally distributed fields of the quantity of grey matter, voxel-wise comparisons can be made over the whole brain in a regression model. Typically the initial matching process involves linear and non-linear warping of the images, which means that voxels need to be stretched and compressed to match up the different subjects. The voxel values can be modulated (scaled) to account for the regional stretching and compression so that the resulting value at each voxel can be considered as a measure of local absolute volume. In the cortex, the local grey matter volume is dependent on local cortical thickness and/or surface area. In healthy brain development, VBM studies have identified grey matter differences associated with normal aging, navigation, arithmetic, linguistic and musical learning abilities (Good et al., 2001; Maguire et al., 2000; Mechelli et al., 2004; Sluming et al., 2002). Further, longitudinal VBM studies have demonstrated training-induced structural changes in the adult human brain (Draganski et al., 2004; Draganski et al., 2006). VBM studies have also identified subtle grey matter differences associated with neurological disorders including diseases with known brain pathology, e.g. Alzheimers disease, Huntington disease (Good et al., 2002; Muhlau et al., 2007) or primary (idiopathic) disorders with the assumption of normal brain morphology, e.g. idiopathic cervical dystonia (Draganski et al., 2003), and chronic pain syndromes (May et al., 1999; Schmidt-Wilcke et al., 2005). Estimation of cortical thickness based on T1-weighted images represents a viable methodological alternative to volumetric measurements for assessment of subtle cortical changes in the human brain. The initial image processing steps required to calculate cortical thickness typically involve segmentation of the images into grey matter, white matter and CSF, as with VBM. However, the subsequent steps may require substantially different processing in order to provide an absolute measure of thickness across the cortical surface. Typically cortical thickness measurements involve identification of the inner and outer cortical boundaries or surfaces. This may be achieved using image information and surface geometry to construct or fit a representation of the grey and white matter surfaces (Fischl and Dale, 2000; Jones et al., 2000; MacDonald et al., 2000; Miller et al., 2000; Zeng et al., 1999). The thickness at each point on the grey matter surface is then given by a distance measure between corresponding points on the two surfaces. Investigating surface-based cortical thickness measures across different subjects requires methods to match corresponding anatomical regions of the cortical surfaces. For example, the FreeSurfer software (http://surfer.nmr.mgh.harvard.edu/, Center for Biomedical Imaging, Charlestown, MA) which has been widely used for measuring cortical thickness in MRI, averages across subjects using non-rigid high-dimensional spherical averaging to align cortical folding patterns (Fischl et al., 1999). Other methods for measuring cortical thickness define the grey and white matter boundaries on the basis of voxel information and the thickness is based on the length of the trajectory from one boundary to another. A distinct difference between these methods and surface-based cortical thickness measures is that thickness is calculated for every volumetric point within the cortex rather than on the surface (Hutton et al., 2008; Jones et al., 2000; Yezzi and Prince, 2003). Differences in cortical thickness have been identified in normal aging (Salat et al., 2004), human intelligence (Choi et al., 2008), cognitive performance (Dickerson et al., 2008), and in neurological disorders including Gilles de la Tourette syndrome in children (Sowell et al., 2008) and Huntington disease (Rosas et al., 2008). Studies using both VBM and surface-based cortical thickness measurements to analyse a single data set have reported differences in the results of the two methods e.g. Blankstein et al. (2009) and Voets et al. (2008). Differences have been attributed both to biology (Voets et al., 2008) and/or methodology (Blankstein et al., 2009). It should be noted that to make robust comparisons between volume- (e.g. VBM) and surface-based anatomical measures (e.g. surface-based cortical thickness), a system is required in which these representations can be integrated e.g. Makris et al. (2006). In this study we compare our voxel-based cortical thickness measure, (VBCT) (Hutton et al., 2008), with the well-established VBM technique (Ashburner and Friston, 2000) in terms of their sensitivity to cortical grey matter changes. The methods are both voxel-based and use the same preprocessing, but measure different features of the grey matter. With VBCT, the one-dimensional scalar thickness of the cortex at each voxel location is measured. Whereas with VBM the quantity of tissue within a voxel is measured which is dependent on the local cortical surface area (and hence cortical folding) as well as the local cortical thickness. Given that both measures are voxel-based, the same spatial registration method can be used so that the comparison can be made conveniently in the same volumetric space. We use a volumetric spatial registration method (DARTEL) which has been demonstrated to provide improved anatomical precision (Ashburner, 2007; Bergouignan et al., 2009; Klein et