Shape prior image segmentation - Caroline Petitjean

Statistical shape prior using manifold learning

In this work, we build a statistical prior using a manifold learning technique (Diffusion Maps) and include it in a graph cut framework. We have described the method in Shape prior based image segmentation using manifold learning, A. Mendoza Quispe and C. Petitjean, accepted at IEEE IPTA, 2015.

Some results on right ventricle segmentation in cardiac magnetic resonance images:

Matlab Code: our segmentation code is available here. The code is compatible with Matlab version 8 with the Image Processing Toolbox.

It requires the use of the following code:

S. Bagon, “Matlab wrapper for graph cuts,” December 2006. http://www.wisdom.weizmann.ac.il/~bagon
L. Van der Maaten, “Matlab toolbox for dimensionality reduction,” http://lvdmaaten.github.io/drtoolbox

Please note: The code is provided without data. The code is a proposition of the implementation of the algorithm that is described in the above-cited paper [Mendoza Quispe and Petitjean, 2015]. It is provided « as is » without any support nor warranty and cannot be guaranteed completely bug-free. We appreciate a citation of our paper in case the code is used.

Main references:

Etyngier, P., Ségonne, F., Keriven, R.: Shape priors using manifold learning techniques. In: Computer Vision, 2007. ICCV 2007. IEEE 11th International Conference on. pp. 18. IEEE (2007)
Moolan-Feroze, O., Mirmehdi, M., Hamilton, M., Bucciarelli-Ducci, C.: Segmentation of the right ventricle using diffusion maps and markov random fields. In: Medical Image Computing and Computer-Assisted InterventionMICCAI 2014, pp. 682689. Springer (2014)

Graph cut image segmentation with multilabel statistical shape prior

Joint work with Damien Grosgeorge and Su Ruan

Matlab code:

to be provided soon

Main references:

D. Grosgeorge, C. Petitjean, S. Ruan, Joint Segmentation of Right and Left Cardiac Ventricles Using Multi-Label Graph Cut, IEEE International Symposium on Biomedical Imaging (ISBI’14), 2014, Beijing, China
D. Grosgeorge, C. Petitjean, J.-N. Dacher, S. Ruan. Graph cut segmentation with a statistical shape model in cardiac MRI, Computer Vision and Image Understanding, vol. 117, p. 1027–1035, 2013.
D. Grosgeorge, C. Petitjean, S. Ruan, J. Caudron, J.-N. Dacher. Right ventricle segmentation by graph cut with shape prior. MICCAI International Workshop, Right Ventricle Segmentation Challenge, 2012.