Multimodal MEMPRAGE, FLAIR, and [Formula: see text] Segmentation to Resolve Dura and Vessels from Cortical Gray Matter.

Roberto Viviani,Daniel Brenner,Julia C Stingl,Eberhard D Pracht,Tony Stöcker,Petra Beschoner

doi:10.3389/fnins.2017.00258

Abstract

While widely in use in automated segmentation approaches for the detection of group differences or of changes associated with continuous predictors in gray matter volume, T1-weighted images are known to represent dura and cortical vessels with signal intensities similar to those of gray matter. By considering multiple signal sources at once, multimodal segmentation approaches may be able to resolve these different tissue classes and address this potential confound. We explored here the simultaneous use of FLAIR and apparent transverse relaxation rates (a signal related to relaxation maps and having similar contrast) with T1-weighted images. Relative to T1-weighted images alone, multimodal segmentation had marked positive effects on 1. the separation of gray matter from dura, 2. the exclusion of vessels from the gray matter compartment, and 3. the contrast with extracerebral connective tissue. While obtainable together with the T1-weighted images without increasing scanning times, apparent transverse relaxation rates were less effective than added FLAIR images in providing the above mentioned advantages. FLAIR images also improved the detection of cortical matter in areas prone to susceptibility artifacts in standard MPRAGE T1-weighted images, while the addition of transverse relaxation maps exacerbated the effect of these artifacts on segmentation. Our results confirm that standard MPRAGE segmentation may overestimate gray matter volume by wrongly assigning vessels and dura to this compartment and show that multimodal approaches may greatly improve the specificity of cortical segmentation. Since multimodal segmentation is easily implemented, these benefits are immediately available to studies focusing on translational applications of structural imaging.

Highlights

Probabilistic tissue classification methods represent one of the most important approaches to the investigation of brain structural differences in vivo with magnetic resonance imaging techniques (Zhang et al, 2001; Fischl et al, 2002, 2004; Ashburner and Friston, 2005)
We explored the use of multimodal segmentation to improve the accuracy of segmentation of cortical gray matter using combinations of MPRAGE, R∗2, and fluid-attenuated inversion recovery (FLAIR) images (Figure 1)
We explored the utility of images of apparent transverse relaxation rates (R∗2 maps), combined with T1-weighted MPRAGE images or in a three-channel combination with MPRAGE and FLAIR

Summary

Introduction

Probabilistic tissue classification methods represent one of the most important approaches to the investigation of brain structural differences in vivo with magnetic resonance imaging techniques (Zhang et al, 2001; Fischl et al, 2002, 2004; Ashburner and Friston, 2005). Unlike work that increases contrast by non-linearly combining images acquired with different modalities into a single image (Misaki et al, 2015), multimodal segmentation considers multiple image types simultaneously and models the intensity of signal from tissue classes as a set of densities in multivariate space. This allows the algorithm, which summarizes the evidence for classification optimally given the density model, to identify the appropriate source of contrast to set the tissue classes apart. We explored the use of multimodal segmentation to improve the accuracy of segmentation of cortical gray matter using combinations of MPRAGE, R∗2, and FLAIR images (Figure 1)

Objectives

Methods

Results

Conclusion