Combined volumetric and surface registration.

Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA 02176 USA.
IEEE transactions on medical imaging 05/2009; 28(4):508-22. DOI: 10.1109/TMI.2008.2004426
Source: PubMed

ABSTRACT In this paper, we propose a novel method for the registration of volumetric images of the brain that optimizes the alignment of both cortical and subcortical structures. In order to achieve this, relevant geometrical information is extracted from a surface-based morph and diffused into the volume using the Navier operator of elasticity, resulting in a volumetric warp that aligns cortical folding patterns. This warp field is then refined with an intensity driven optical flow procedure that registers noncortical regions, while preserving the cortical alignment. The result is a combined surface and volume morph (CVS) that accurately registers both cortical and subcortical regions, establishing a single coordinate system suitable for the entire brain.

1 Bookmark
  • [Show abstract] [Hide abstract]
    ABSTRACT: Whole brain mapping of stroke patients with large cortical infarcts is not trivial due to the complexity of infarcts' anatomical location and appearance in magnetic resonance image. In this study, we proposed an individualized diffeomorphic mapping framework for solving this problem. This framework is based on our recent work of large deformation diffeomorphic metric mapping (LDDMM) in Du et al. (2011) and incorporates anatomical features, such as sulcal/gyral curves, cortical surfaces, brain intensity image, and masks of infarcted regions, in order to align a normal brain to the brain of stroke patients. We applied this framework to synthetic data and data of stroke patients and validated the mapping accuracy in terms of the alignment of gyral/sulcal curves, sulcal regions, and brain segmentation. Our results revealed that this framework provided comparable mapping results for stroke patients and healthy controls, suggesting the importance of incorporating individualized anatomical features in whole brain mapping of brains with large cortical infarcts.
    Magnetic Resonance Imaging 09/2014; · 2.02 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The position of cortical areas can be approximately predicted from cortical surface folding patterns. However, there is extensive inter-subject variability in cortical folding patterns, prohibiting a one-to-one mapping of cortical folds in certain areas. In addition, the relationship between cortical area boundaries and the shape of the cortex is variable, and weaker for higher-order cortical areas. Current surface registration techniques align cortical folding patterns using sulcal landmarks or cortical curvature, for instance. The alignment of cortical areas by these techniques is thus inherently limited by the sole use of geometric similarity metrics. Magnetic resonance imaging T1 maps show intra-cortical contrast that reflects myelin content, and thus can be used to improve the alignment of cortical areas. In this article, we present a new symmetric diffeomorphic multi-contrast multi-scale surface registration (MMSR) technique that works with partially inflated surfaces in the level-set framework. MMSR generates a more precise alignment of cortical surface curvature in comparison to two widely recognized surface registration algorithms. The resulting overlap in gyri labels is comparable to FreeSurfer. Most importantly, MMSR improves the alignment of cortical areas further by including T1 maps. As a first application, we present a group average T1 map at a uniquely high-resolution and multiple cortical depths, which reflects the myeloarchitecture of the cortex. MMSR can also be applied to other MR contrasts, such as functional and connectivity data. Copyright © 2015. Published by Elsevier Inc.
    NeuroImage 02/2015; 11. · 6.13 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we present a novel label fusion algorithm suited for scenarios in which different manual delineation protocols with potentially disparate structures have been used to annotate the training scans (hereafter referred to as “atlases”). Such scenarios arise when atlases have missing structures, when they have been labeled with different levels of detail, or when they have been taken from different heterogeneous databases. The proposed algorithm can be used to automatically label a novel scan with any of the protocols from the training data. Further, it enables us to generate new labels that are not present in any delineation protocol by defining intersections on the underling labels. We first use probabilistic models of label fusion to generalize three popular label fusion techniques to the multi-protocol setting: majority voting, semi-locally weighted voting and STAPLE. Then, we identify some shortcomings of the generalized methods, namely the inability to produce meaningful posterior probabilities for the different labels (majority voting, semi-locally weighted voting) and to exploit the similarities between the atlases (all three methods). Finally, we propose a novel generative label fusion model that can overcome these drawbacks. We use the proposed method to combine four brain MRI datasets labeled with different protocols (with a total of 102 unique labeled structures) to produce segmentations of 148 brain regions. Using cross-validation, we show that the proposed algorithm outperforms the generalizations of majority voting, semi-locally weighted voting and STAPLE (mean Dice score 83%, vs. 77%, 80% and 79%, respectively). We also evaluated the proposed algorithm in an aging study, successfully reproducing some well-known results in cortical and subcortical structures.
    NeuroImage 11/2014; 106. · 6.13 Impact Factor


Available from