Anomaly detection through registration

Abstract We introduce a system that automatically segments and classifies features in brain MRI volumes. It segments 144 structures of a 256×256×124 voxel image in 18 minutes on an SGI computer with four 194 MHz R10K processors. The algorithm uses an atlas, a hand-segmented and classified MRI of a normal brain, which is warped in 3-D using a hierarchical deformable matching algorithm until it closely matches the subject. This customized atlas contains the segmentation and classification of the subject’s anatomical structures. We have conducted tests on 139 MRIs of normal brains, and 3 MRIs and 1 CT of brains with pathologies. We present qualitative and quantitative evaluations of the system’s performance. Combined with domain knowledge, the registration algorithm is capable of detecting asymmetries and abnormal variations in the subject’s data that indicate the existence and location of pathologies.

[1]  Max A. Viergever,et al.  Evaluation of Ridge Seeking Operators for Multimodality Medical Image Matching , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Paul A. Viola,et al.  Multi-modal volume registration by maximization of mutual information , 1996, Medical Image Anal..

[3]  Michael I. Miller,et al.  Deformable templates using large deformation kinematics , 1996, IEEE Trans. Image Process..

[4]  Philippe Cinquin,et al.  Image guided operating robot: a clinical application in stereotactic neurosurgery , 1992, Proceedings 1992 IEEE International Conference on Robotics and Automation.

[5]  R. Bajcsy,et al.  Elastically Deforming 3D Atlas to Match Anatomical Brain Images , 1993, Journal of computer assisted tomography.

[6]  Guido Gerig,et al.  Segmentation of 2-D and 3-D objects from MRI volume data using constrained elastic deformations of flexible Fourier contour and surface models , 1996, Medical Image Anal..

[7]  Berthold K. P. Horn,et al.  Closed-form solution of absolute orientation using unit quaternions , 1987 .

[8]  Stan Sclaroff,et al.  On modal modeling for medical images: underconstrained shape description and data compression , 1994, Proceedings of IEEE Workshop on Biomedical Image Analysis.

[9]  M. Viergever,et al.  Medical image matching-a review with classification , 1993, IEEE Engineering in Medicine and Biology Magazine.

[10]  Alex Pentland,et al.  Modal Matching for Correspondence and Recognition , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[11]  Jean-Philippe Thirion,et al.  Fast Non-Rigid Matching of 3D Medical Images , 1995 .

[12]  Paul A. Viola,et al.  Alignment by Maximization of Mutual Information , 1995, Proceedings of IEEE International Conference on Computer Vision.

[13]  Alex Pentland,et al.  A Bayesian similarity measure for direct image matching , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[14]  Michael I. Miller,et al.  Individualizing Neuroanatomic Atlases Using a Massively Parallel Computer , 1996, Computer.

[15]  Ruzena Bajcsy,et al.  Multiresolution elastic matching , 1989, Comput. Vis. Graph. Image Process..

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

[17]  Guy Marchal,et al.  Multimodality image registration by maximization of mutual information , 1997, IEEE Transactions on Medical Imaging.