A Bayesian model for joint segmentation and registration

By Kilian M. Pohl, John W. Fisher III, W. Grimson, Ron Kikinis, William M. Wells. NeuroImage, 2006

312 citations recorded by the source corpus.

A statistical model is presented that combines the registration of an atlas with the segmentation of magnetic resonance images. We use an Expectation Maximization-based algorithm to find a solution within the model, which simultaneously estimates image artifacts, anatomical labelmaps, and a structure-dependent hierarchical mapping from the atlas to the image space. The algorithm produces segmentations for brain tissues as well as their substructures. We demonstrate the approach on a set of 22 magnetic resonance images. On this set of images, the new approach performs significantly better than similar methods which sequentially apply registration and segmentation.

Subfields: Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/98033f115fe7df271868eb855aebf3d2935a68ec

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

DRAMMS: Deformable Registration via Attribute Matching and Mutual-Saliency Weighting (2009, background): “Other works explicitly segment the regions of interest (ROI) in a joint segmentation and registration framework (Yezzi et al., 2001; Wyatt and Nobel, 2003; Chen et al., 2005; Pohl et al., 2006; Wang et al., 2006; Xue et al., 2008; Ou et al., 2009a; Periaswamy and Farid, 2006) and assign higher weigh”
DeepNAT: Deep Convolutional Neural Network for Segmenting Neuroanatomy (2017, background): “multi-atlas techniques and label fusion strategies (Ashburner and Friston, 2005; Corresponding Author. Address: Waltherstr. 23, 81369 Munc hen, Germany; Email: christian.wachinger@med.unimuenchen.de Pohl et al., 2006; Heckemann et al., 2006; Rohlng et al., 2004, 2005; Svarer et al., 2005; Sabuncu et”
MRI Segmentation of the Human Brain: Challenges, Methods, and Applications (2015, methodology): “Bayesian classifiers are used in the expectation-maximization (EM) segmentation methods which have been successfully implemented in several software packages used in the medical imaging community: SPM [2], FAST [3], FreeSurfer [21], and 3DSlicer [71].”
Brain Functional Localization: A Survey of Image Registration Techniques (2007, unknown): “Validation through manual expert segmentations has been presented in several articles [63], [251], [258].”
A review of atlas-based segmentation for magnetic resonance brain images (2011, methodology): “This behaviour can also be observed in Table 3, where the best Dice values for each structure are obtained from multi-atlas approaches [65–67], closely followed by statistical frameworks [55,78,108].”
Fast and robust multi-atlas segmentation of brain magnetic resonance images (2010, methodology): “Many different approaches have been published using registrationbased segmentation, for example, for segmenting subcortical structures (Avants et al., 2008; Bhattacharjee et al., 2008; Han and Fischl, 2007; Pohl et al., 2006).”
A Bayesian model of shape and appearance for subcortical brain segmentation (2011, unknown): “The most common volumetric-based approaches to segmentation are based on non-linear warping of an atlas, or atlases, to new data (Collins and Evans, 1997; Fischl et al., 2002; Pohl et al., 2006).”
A probabilistic atlas of the human thalamic nuclei combining ex vivo MRI and histology (2018, methodology): “…Saygin et al., 2017), and following the literature of Bayesian segmentation (Wells et al., 1996; Zhang et al., 2001; Ashburner and Friston, 2005; Pohl et al., 2006), we assumed the following forward model: first, the atlas is spatially warped following a deformation model (Ashburner et al.,…”
A dynamic 4D probabilistic atlas of the developing brain (2011, background): “State-of-the-art techniques for automatic brain MRI segmentation typically depend on spatial prior information in form of a probabilistic atlas (Ashburner and Friston, 2005; Pohl et al., 2006).”
A Review of Geometric Transformations for Nonrigid Body Registration (2008, background): “Atlases can be either probabilistic [33]–[35], intensity based [36], label based [37], [38], or deformation based [39].”
Multi-spectral brain tissue segmentation using automatically trained k-Nearest-Neighbor classification (2007, methodology): “Ashburner and Friston (2005), D'Agostino et al. (2006) and Pohl et al. (2006) extended this work by incorporating the registration step in the EM algorithm, thus simultaneously estimating tissue labels, and the mapping from the atlas to the image space.”
Multi-Atlas Segmentation of Biomedical Images: A Survey (2014, background): “However, the approach that dominated early atlas-guided segmentation was probabilistic atlas-based segmentation (Ashburner and Friston, 2005; Fischl et al., 2002; Park et al., 2003; Pohl et al., 2006; Thomas Yeo et al., 2008), which had two distinctive properties.”
Automated segmentation of hippocampal subfields from ultra‐high resolution in vivo MRI (2009, methodology): “Optimizing the lower bound with respect to x reduces to optimizing and thus involves deforming the atlas towards the current clas-siﬁcation, similar to the schemes proposed in (Pohl et al., 2006; D’Agostino et al., 2006).”
An Open Source Multivariate Framework for n-Tissue Segmentation with Evaluation on Public Data (2011, methodology): “Similar integrative brain processing was provided in [33] in which segmentation and registration parameters were optimized simultaneously while casting the inhomogeneity model parameters of [15] as nuisance variables.”
Auto-Context and Its Application to High-Level Vision Tasks and 3D Brain Image Segmentation (2010, background): “These algorithms range from shape-driven [54], [30], atlas and knowledge-based [35], Markov Random Fields models [10], [31], to classification/learning-based approaches [24].”