Learning Correspondences in Knee MR Images from the Osteoarthritis Initiative

Registration is a powerful tool that allows mapping images in a common space in order to aid in their analysis. Accurate registration of images of the knee is challenging to achieve using intensity based registration algorithms. Problems arise due to large anatomical inter-subject differences which causes registrations to fail to converge to an accurate solution. In this work we propose learning correspondences in pairs of images to match self-similarity features, that describe images in terms of their local structure rather than their intensity. We use RANSAC as a robust model estimator. We show a substantial improvement in terms of mean error and standard deviation of 2.13mm and 2.47mm over intensity based registration methods, when comparing landmark alignment error.

[1]  Pascal Fua,et al.  A parallel stereo algorithm that produces dense depth maps and preserves image features , 1993, Machine Vision and Applications.

[2]  Daniel Rueckert,et al.  Nonrigid registration using free-form deformations: application to breast MR images , 1999, IEEE Transactions on Medical Imaging.

[3]  Sharmila Majumdar,et al.  Atlas‐based knee cartilage assessment , 2011, Magnetic resonance in medicine.

[4]  Jürgen Weese,et al.  Landmark-based elastic registration using approximating thin-plate splines , 2001, IEEE Transactions on Medical Imaging.

[5]  Eli Shechtman,et al.  Matching Local Self-Similarities across Images and Videos , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[6]  Christos Davatzikos,et al.  GRAM: A framework for geodesic registration on anatomical manifolds , 2010, Medical Image Anal..

[7]  Daniel Rueckert,et al.  Multi-atlas based segmentation of brain images: Atlas selection and its effect on accuracy , 2009, NeuroImage.

[8]  Daniel Rueckert,et al.  Manifold learning for automatically predicting articular cartilage morphology in the knee with data from the osteoarthritis initiative (OAI) , 2011, Medical Imaging.

[9]  Karl Rohr.  On 3D differential operators for detecting point landmarks , 1997, Image Vis. Comput..

[10]  Patrik O. Hoyer,et al.  Non-negative Matrix Factorization with Sparseness Constraints , 2004, J. Mach. Learn. Res..

[11]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[12]  William M. Wells,et al.  Medical Image Computing and Computer-Assisted Intervention — MICCAI’98 , 1998, Lecture Notes in Computer Science.

[13]  Xavier Pennec,et al.  Feature-Based Registration of Medical Images: Estimation and Validation of the Pose Accuracy , 1998, MICCAI.

[14]  Erika Schneider,et al.  The osteoarthritis initiative: report on the design rationale for the magnetic resonance imaging protocol for the knee. , 2008, Osteoarthritis and cartilage.

[15]  Timothy F. Cootes,et al.  The Use of Active Shape Models for Locating Structures in Medical Images , 1993, IPMI.

[16]  Dinggang Shen,et al.  Directed graph based image registration , 2012, Comput. Medical Imaging Graph..

[17]  Daniel Rueckert,et al.  Landmark localisation in brain MR images using feature point descriptors based on 3D local self-similarities , 2012, 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI).