Generation of a statistical shape model with probabilistic point correspondences and the expectation maximization- iterative closest point algorithm

Objective:Identification of point correspondences between shapes is required for statistical analysis of organ shapes differences. Since manual identification of landmarks is not a feasible option in 3D, several methods were developed to automatically find one-to-one correspondences on shape surfaces. For unstructured point sets, however, one-to-one correspondences do not exist but correspondence probabilities can be determined.Materials and methods: A method was developed to compute a statistical shape model based on shapes which are represented by unstructured point sets with arbitrary point numbers. A fundamental problem when computing statistical shape models is the determination of correspondences between the points of the shape observations of the training data set. In the absence of landmarks, exact correspondences can only be determined between continuous surfaces, not between unstructured point sets. To overcome this problem, we introduce correspondence probabilities instead of exact correspondences. The correspondence probabilities are found by aligning the observation shapes with the affine expectation maximization-iterative closest points (EM-ICP) registration algorithm. In a second step, the correspondence probabilities are used as input to compute a mean shape (represented once again by an unstructured point set). Both steps are unified in a single optimization criterion which depe nds on the two parameters ‘registration transformation’ and ‘mean shape’. In a last step, a variability model which best represents the variability in the training data set is computed. Experiments on synthetic data sets and in vivo brain structure data sets (MRI) are then designed to evaluate the performance of our algorithm.Results:The new method was applied to brain MRI data sets, and the estimated point correspondences were compared to a statistical shape model built on exact correspondences. Based on established measures of “generalization ability” and “specificity”, the estimates were very satisfactory.Conclusion:The novel algorithm for building a generative statistical shape model (gSSM) does not need one-to-one point correspondences but relies solely on point correspondence probabilities for the computation of mean shape and eigenmodes. It is well-suited for shape analysis on unstructured point sets.

[1]  Timothy F. Cootes,et al.  Active Shape Models-Their Training and Application , 1995, Comput. Vis. Image Underst..

[2]  Robert T. Schultz,et al.  A unified non-rigid feature registration method for brain mapping , 2003, Medical Image Anal..

[3]  E LorensenWilliam,et al.  Decimation of triangle meshes , 1992 .

[4]  Anand Rangarajan,et al.  Unsupervised learning of an Atlas from unlabeled point-sets , 2004, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[5]  Anand Rangarajan,et al.  The Softassign Procrustes Matching Algorithm , 1997, IPMI.

[6]  William E. Lorensen,et al.  Decimation of triangle meshes , 1992, SIGGRAPH.

[7]  F. Bookstein Landmark methods for forms without landmarks , 1996 .

[8]  Anand Rangarajan,et al.  Kernel Fisher discriminant for shape-based classification in epilepsy , 2007, Medical Image Anal..

[9]  William M. Wells,et al.  An EM algorithm for shape classification based on level sets , 2005, Medical Image Anal..

[10]  Sébastien Granger,et al.  Une approche statistique multi-échelle au recalage rigide de surfaces : Application à l'implantologie dentaire , 2003 .

[11]  Hans-Peter Meinzer,et al.  3D Active Shape Models Using Gradient Descent Optimization of Description Length , 2005, IPMI.

[12]  Timothy F. Cootes,et al.  A minimum description length approach to statistical shape modeling , 2002, IEEE Transactions on Medical Imaging.

[13]  Xavier Pennec,et al.  Multi-scale EM-ICP: A Fast and Robust Approach for Surface Registration , 2002, ECCV.

[14]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[15]  Martin Styner,et al.  Statistical shape analysis of neuroanatomical structures based on medial models , 2003, Medical Image Anal..

[16]  Nicholas Ayache,et al.  Medical Image Analysis: Progress over Two Decades and the Challenges Ahead , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[17]  Paul J. Besl,et al.  A Method for Registration of 3-D Shapes , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Miriah D. Meyer,et al.  Entropy-Based Particle Systems for Shape Correspondence , 2006 .

[19]  Cristian Lorenz,et al.  Generation of Point-Based 3D Statistical Shape Models for Anatomical Objects , 2000, Comput. Vis. Image Underst..

[20]  Frans Vos,et al.  A statistical shape model without using landmarks , 2004, Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004..

[21]  Martin Styner,et al.  Evaluation of 3D Correspondence Methods for Model Building , 2003, IPMI.

[22]  Anand Rangarajan,et al.  Shape analysis using the Fisher-Rao Riemannian metric: unifying shape representation and deformation , 2006, 3rd IEEE International Symposium on Biomedical Imaging: Nano to Macro, 2006..

[23]  Eam Khwang Teoh,et al.  A novel framework for automated 3D PDM construction using deformable models , 2005, SPIE Medical Imaging.