Efficient Use of Cerebral Cortical Thickness to Correct Brain MR Segmentation

Abstract

Efficient, automatic and robust tools for measurement of cerebral cortical thickness would aid diagnosis and longitudinal studies of neurodegenerative disorders. In this work, we segment a 3D magnetic resonance image of the brain using an Expectation-Maximization approach. The definition of thickness used is based on the solution of Laplace's equation in the cortex. Unlike other works, finite difference equations for calculation of cortical thickness are generalized for anisotropic images in order to avoid resampling the input images. We also developed a method which combines information from the thickness estimation with the segmentation probability maps, in order to detect missegmented sulci and correct the segmentation accordingly.

A comparison of voxel and surface based cortical thickness estimation methods

“…Laplacian approaches (Hutton et al, 2008;Acosta et al, 2009;Cardoso et al, 2011), solve the Laplace equation (Jones et al, 2000) using boundary value relaxation (Press et al, 1991) or matrix methods (Haidar et al, 2005), calculate thickness by integrating the tangent to the Laplacian scalar field (Jones et al, 2000), summing the Euclidean distance from neighbouring voxels on the same streamline, or using a partial differential equation (Yezzi and Prince, 2003) with boundaries set to zero (Yezzi and Prince, 2003), half the mean voxel dimension (Diep et al, 2007) or using Lagrangian initialisation (Bourgeat et al, 2008;Acosta et al, 2009). In contrast, the registration based approach of Das et al (2009) uses a greedy diffeomorphic registration algorithm to warp the WM segment to match the GM+WM segment.…”

Analysis of Cortical Thickness Indicating Cingulate Gyrus and Temporal Gyrus Incrassation in Posttraumatic Stress Disorder Due to Mining Disaster

“…However, due to noise, intensity inhomogeneities, and partial volume effect, it is hard to classify the tissues perfectly, especially in the sulci regions. In this study, we used the algorithm proposed by Thanh-Mai Diep [16] to further improve the brain segmentation. Sulci detection was performed after the thickness estimation algorithm.…”

“…However, due to limited resolution of MRI and the complexity of the 3D curve structure, erroneously high thickness estimates may be obtained for some regions. Considering the importance of accurate segmentation of gray matter on thickness estimation, mixturebased segmentation with probability maps was used to get accurate detection of the cortex, especially that for deep sulci, which in turn improves the accuracy of the thickness estimation [16]. In this study, to compare the cortical thickness between survivors of mining disasters with PTSD and survivors without PTSD, statistical parametric mapping (SPM) was applied and the regions of anterior cingulated cortex, superior temporal gyrus, and transverse temporal gyrus were studied.…”

Automatic Delineation of Sulci and Improved Partial Volume Classification for Accurate 3D Voxel-Based Cortical Thickness Estimation from MR

“…The approaches previously proposed in the literature can be broadly categorised as mesh-based [6] and voxel-based [1,2,3,4,5]. Operating directly on the 3D voxel grid, voxel based techniques are more computationally efficient but less robust to noise and missegmentation as they typically lack the mechanisms required to assess and correct topological errors in gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF) segmentations.…”

“…However, no additional information is used to ensure that sulci are correctly identified and does not account for partial volume information to accurately compute the thickness. Diep [4] proposed to cut sulci using the probability of the voxel being CSF in abnormally thick areas, but it lacks a partial volume model to properly identify CSF voxels buried inside deep sulci.…”