Image Registration by Hierarchical Matching of Local Spatial Intensity Histograms

We previously presented a HAMMER image registration algorithm that demonstrated high accuracy in superposition of images from different individual brains. However, the HAMMER registration algorithm requires pre-segmentation of brain tissues, since the attribute vectors used to hierarchically match the corresponding pairs of points are defined from the segmented images. In many applications, the segmentation of tissues might be difficult, unreliable or even impossible to complete, which potentially limits the use of the HAMMER algorithm in more generalized applications. To overcome this limitation, we use local spatial intensity histograms to design a new type of attribute vector for each point in an intensity image. The histogram-based attribute vector is rotationally invariant, and more importantly it captures spatial information by integrating a number of local histograms that are calculated from multi-resolution images. The new attribute vectors are able to determine corresponding points across individual images. Therefore, by hierarchically matching new attribute vectors, the proposed registration method performs as successfully as the previous HAMMER algorithm did in registering MR brain images, while providing more general applications in registering images of other organs. Experimental results show good performance of the proposed method in registering MR brain images and CT pelvis images.

[1]  Shree K. Nayar,et al.  Spatial information in multiresolution histograms , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[2]  Ruzena Bajcsy,et al.  Matching structural images of the human brain using statistical and geometrical image features , 1994, Other Conferences.

[3]  Dinggang Shen,et al.  Generalized Affine Invariant Image Normalization , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  Dinggang Shen,et al.  HAMMER: hierarchical attribute matching mechanism for elastic registration , 2002, IEEE Transactions on Medical Imaging.

[5]  Gary E. Christensen,et al.  Consistent image registration , 2001, IEEE Transactions on Medical Imaging.

[6]  Robert T. Schultz,et al.  A Unified Feature Registration Method for Brain Mapping , 2001, IPMI.

[7]  Dinggang Shen,et al.  Very High-Resolution Morphometry Using Mass-Preserving Deformations and HAMMER Elastic Registration , 2003, NeuroImage.

[8]  Michael I. Miller,et al.  Hierarchical brain mapping via a generalized Dirichlet solution for mapping brain manifolds , 1995, Optics & Photonics.

[9]  Demetri Terzopoulos,et al.  Deformable models in medical image analysis: a survey , 1996, Medical Image Anal..

[10]  Paul M. Thompson,et al.  A surface-based technique for warping three-dimensional images of the brain , 1996, IEEE Trans. Medical Imaging.

[11]  D. Louis Collins,et al.  Warping of a computerized 3-D atlas to match brain image volumes for quantitative neuroanatomical and functional analysis , 1991, Medical Imaging.

[12]  Nicholas Ayache,et al.  Automatic registration of 3D images using surface curvature , 1992, Optics & Photonics.

[13]  S. Resnick,et al.  One-year age changes in MRI brain volumes in older adults. , 2000, Cerebral cortex.

[14]  John F. Canny,et al.  A Computational Approach to Edge Detection , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  Dinggang Shen,et al.  Correspondence Detection Using Wavelet-Based Attribute Vectors , 2003, MICCAI.