Higher-Order CRF Tumor Segmentation with Discriminant Manifold Potentials

The delineation of tumor boundaries in medical images is an essential task for the early detection, diagnosis and follow-up of cancer. However accurate segmentation remains challenging due to presence of noise, inhomogeneity and high appearance variability of malignant tissue. In this paper, we propose an automatic segmentation approach using fully-connected higher-order conditional random fields (HOCRF) where potentials are computed within a discriminant Grassmannian manifold. First, the framework learns within-class and between-class similarity distributions from a training set of images to discover the optimal manifold discrimination between normal and pathological tissues. Second, the conditional optimization scheme computes non-local pairwise as well as pattern-based higher-order potentials from the manifold subspace to recognize regions with similar labelings and incorporate global consistency in the inference process. Our HOCRF framework is applied in the context of metastatic liver tumor segmentation in CT images. Compared to state of the art methods, our method achieves better performance on a group of 30 liver tumors and can deal with highly pathological cases.

[1]  Matthieu Guillaumin,et al.  Segmentation Propagation in ImageNet , 2012, ECCV.

[2]  Stefan Bauer,et al.  Fully Automatic Segmentation of Brain Tumor Images Using Support Vector Machine Classification in Combination with Hierarchical Conditional Random Field Regularization , 2011, MICCAI.

[3]  Vibhav Vineet,et al.  Filter-Based Mean-Field Inference for Random Fields with Higher-Order Terms and Product Label-Spaces , 2012, International Journal of Computer Vision.

[4]  S T Roweis,et al.  Nonlinear dimensionality reduction by locally linear embedding. , 2000, Science.

[5]  Mark Jenkinson,et al.  Non-local Shape Descriptor: A New Similarity Metric for Deformable Multi-modal Registration , 2011, MICCAI.

[6]  Andrew McCallum,et al.  Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data , 2001, ICML.

[7]  Chunming Li,et al.  Active contours driven by local Gaussian distribution fitting energy , 2009, Signal Process..

[8]  Stan Z. Li,et al.  Markov Random Field Modeling in Image Analysis , 2001, Computer Science Workbench.

[9]  Vladlen Koltun,et al.  Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials , 2011, NIPS.

[10]  Daniel Rueckert,et al.  Hierarchical Manifold Learning , 2012, MICCAI.

[11]  Greg Mori,et al.  Guiding model search using segmentation , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[12]  Pushmeet Kohli,et al.  Robust Higher Order Potentials for Enforcing Label Consistency , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[13]  Hervé Delingette,et al.  Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012 , 2012, Lecture Notes in Computer Science.

[14]  Mika Pollari,et al.  Semi-automatic liver tumor segmentation with hidden Markov measure field model and non-parametric distribution estimation , 2012, Medical Image Anal..

[15]  Brian C. Lovell,et al.  Graph embedding discriminant analysis on Grassmannian manifolds for improved image set matching , 2011, CVPR 2011.

[16]  Mert R. Sabuncu,et al.  A Generative Model for Image Segmentation Based on Label Fusion , 2010, IEEE Transactions on Medical Imaging.