Region-Specific Hierarchical Segmentation of MR Prostate Using Discriminative Learning

We demonstrate the effectiveness of learning-based methods and hierarchical boundary deformation for efficient, accurate segmentation of prostate in T2 weighted MRI data. After normalizing intraand inter-image intensity variation, Marginal Space Learning (MSL) is used to align a statistical mesh model on the image. This mesh is then hierarchically refined to the image boundary using spatially varying surface classifiers. Using 10-fold cross validation on 50 cases, we achieve accurate delineations (0.89 dice coeff., 1.91mm surface error) in under 3s.

[1]  Jocelyne Troccaz,et al.  Atlas-based prostate segmentation using an hybrid registration , 2008, International Journal of Computer Assisted Radiology and Surgery.

[2]  Filip Claus,et al.  Interobserver Delineation Variation Using CT versus Combined CT + MRI in Intensity–Modulated Radiotherapy for Prostate Cancer , 2005, Strahlentherapie und Onkologie.

[3]  Steve Webb,et al.  Radiotherapy treatment planning of prostate cancer using magnetic resonance imaging alone. , 2003, Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology.

[4]  Gabor Fichtinger,et al.  Prostate contouring in MRI guided biopsy , 2009, Medical Imaging.

[5]  Dorin Comaniciu,et al.  Hierarchical, learning-based automatic liver segmentation , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[6]  Dorin Comaniciu,et al.  Four-Chamber Heart Modeling and Automatic Segmentation for 3-D Cardiac CT Volumes Using Marginal Space Learning and Steerable Features , 2008, IEEE Transactions on Medical Imaging.

[7]  Shaohua Kevin Zhou,et al.  Integrated Detection Network (IDN) for pose and boundary estimation in medical images , 2011, 2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[8]  M van Herk,et al.  Definition of the prostate in CT and MRI: a multi-observer study. , 1999, International journal of radiation oncology, biology, physics.

[9]  Zhuowen Tu,et al.  Probabilistic boosting-tree: learning discriminative models for classification, recognition, and clustering , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[10]  Stefan Klein,et al.  Automatic segmentation of the prostate in 3D MR images by atlas matching using localized mutual information. , 2008, Medical physics.

[11]  Jocelyne Troccaz,et al.  Automated segmentation of the prostate in 3D MR images using a probabilistic atlas and a spatially constrained deformable model. , 2010, Medical physics.

[12]  Patrick Pérez,et al.  Poisson image editing , 2003, ACM Trans. Graph..