Label fusion for segmentation via patch based on local weighted voting

Abstract

Label fusion is a powerful image segmentation strategy that is becoming increasingly popular in medical imaging. However, satisfying the requirements of higher accuracy and less running time is always a great challenge. In this paper we propose a novel patch-based segmentation method combining a local weighted voting strategy with Bayesian inference. Multiple atlases are registered to a target image by an advanced normalization tools (ANTs) algorithm. To obtain a segmentation of the target, labels of the atlas images are propagated to the target image. We first adopt intensity prior and label prior as two key metrics when implementing the local weighted voting scheme, and then compute the two priors at the patch level. Further, we analyze the label fusion procedure concerning the image background and take the image background as an isolated label when estimating the label prior. Finally, by taking the Dice score as a criterion to quantitatively assess the accuracy of segmentations, we compare the results with those of other methods, including joint fusion, majority voting, local weighted voting, majority voting based on patch, and the widely used FreeSurfer whole-brain segmentation tool. It can be clearly seen that the proposed algorithm provides better results than the other methods. During the experiments, we make explorations about the influence of different parameters (including patch size, patch area, and the number of training subjects) on segmentation accuracy.

Keywords

Label fusion / Local weighted voting / Patch-based / Background analysis

References

[1] Asman,A.J., Landman, B.A., 2013. Non-local statistical label fusion for multi-atlas segmentation. Med. Image Anal., 17(2):194–208. http://dx.doi.org/10.1016/j.media.2012.10.002

[2] Avants,B.B., Tustison, N.J., Song,G. , , 2011. A reproducible evaluation of ANTs similarity metric performance in brain image registration. NeuroImage, 54(3):2033–2044. http://dx.doi.org/10.1016/j.neuroimage.2010.09.025

[3] Cocosco,C., Zijdenbos, A., Evans,A. , 2003. A fully automatic and robust brain MRI tissue classification method. Med. Image Anal., 7(4):513–527. http://dx.doi.org/10.1016/S1361-8415(03)00037-9

[4] Coupé,P., Manjón, J., Fonov,V. , , 2011. Patch-based segmentation using expert priors: application to hippocampus and ventricle segmentation. NeuroImage, 54(2): 940–954. http://dx.doi.org/10.1016/j.neuroimage.2010.09.018

[5] Coupé,P., Eskildsen, S.F., Manjón,J.V. , , 2012. Simultaneous segmentation and grading of anatomical structures for patient’s classification: application to Alzheimer’s disease. NeuroImage, 59(4):3736–3747. http://dx.doi.org/10.1016/j.neuroimage.2011.10.080

[6] Eskildsen,S.F., Coupé, P., Fonov,V. , , 2012. BEaST: brain extraction based on nonlocal segmentation technique. NeuroImage, 59(3):2362–2373. http://dx.doi.org/10.1016/j.neuroimage.2011.09.012

[7] Ghasemi,J., Mollaei, M.R.K., Ghaderi,R. , , 2012. Brain tissue segmentation based on spatial information fusion by Dempster-Shafer theory. J. Zhejiang Univ.-Sci. C (Comput. & Electron.), 13(7):520–533. http://dx.doi.org/10.1631/jzus.C1100288

[8] Isgum,I., Staring, M., Rutten,A. , , 2009. Multi-atlasbased segmentation with local decision fusion—application to cardiac and aortic segmentation in CT scans. IEEE Trans. Med. Imag., 28(7):1000–1010. http://dx.doi.org/10.1109/TMI.2008.2011480

[9] Manjón,J.V., Carbonell-Caballero, J., Lull,J.J. ,, 2008. MRI denoising using non-local means. Med. Image Anal., 12(4):514–523. http://dx.doi.org/10.1016/j.media.2008.02.004

[10] Rohlfing,T., Brandt, R., Menzel,R. , , 2004. Evaluation of atlas selection strategies for atlas-based image segmentation with application to confocal microscopy images of bee brains. NeuroImage, 21(4):1428–1442. http://dx.doi.org/10.1016/j.neuroimage.2003.11.010

[11] Rousseau,F., Habas, P.A., Studholme,C. , 2011. A supervised patch-based approach for human brain labeling. IEEE Trans. Med. Imag., 30(10):1852–1862. http://dx.doi.org/10.1109/TMI.2011.2156806

[12] Sabuncu,M.R., Yeo,B.T.T., van Leemput,K. , , 2010. A generative model for image segmentation based on label fusion. IEEE Trans. Med. Imag., 29(10):1714–1729. http://dx.doi.org/10.1109/TMI.2010.2050897

[13] Shi,F., Yang,J., Zhu,Y.M., 2009. Automatic segmentation of bladder in CT images. J. Zhejiang Univ.-Sci. A, 10(2): 239–246. http://dx.doi.org/10.1631/jzus.A0820157

[14] Wang,H.Z., Yushkevich, P.A., 2013. Multi-atlas segmentation with joint label fusion and corrective learning—an open source implementation. Front. Neuroinform., 7:27. http://dx.doi.org/10.3389/fninf.2013.00027

[15] Wang,H.Z., Suh,J.W., Das,S.R., , 2013. Multi-atlas segmentation with joint label fusion. IEEE Trans. Patt. Anal. Mach. Intell., 35(3):611–623. http://dx.doi.org/10.1109/TPAMI.2012.143

[16] Wang,L., Shi,F., Li,G., , 2014. Segmentation of neonatal brain MR images using patch-driven level sets. NeuroImage, 84(1):141–158. http://dx.doi.org/10.1016/j.neuroimage.2013.08.008

RIGHTS & PERMISSIONS

2017 Zhejiang University and Springer-Verlag Berlin Heidelberg