___

• have applied the MLP-based liver segmentation method after several preprocessing steps and obtained more
• successful results. However, the drawback of this method is a very high computational cost.
• We also have explained the properties of two different recently published level set-based segmentation
• techniques and presented their implementation results for liver segmentation. The FTC algorithm, which uses
• a switching mechanism, seems successful and can give acceptable results when preprocessed abdominal images
• are used. However, the drawback of the FTC algorithm is its sensitivity for initial contours that are defined
• in each slice. This is because the accuracy of the results depends not only on the size of the initial contours
• drawn by users, but also on the number of initial contours and their positions. In addition, the user-defined
• iteration numbers for each slice affect the segmentation results. Therefore, this approach is not robust, and it
• generates oversegmented or undersegmented images on some slices. In order to overcome these drawbacks, we
• have applied an automatic method iteratively using the FTC algorithm without any user interaction and with
• a small fixed number of iterations for liver segmentation from SPIR datasets. However, we have observed that
• automatic liver segmentation using the FTC method is not successful for SPIR datasets, even if preprocessed
• images are used. Therefore, we have applied the DRLSE with SPF function-based automatic segmentation
• method, which gives fast and acceptable results without any postprocessing operation. Moreover, this method
• has the ability to segment liver images without extracting adjacent organs except the gallbladder. In addition,
• it is more efficient than the MLP-based segmentation method in terms of the required segmentation time.
• Both qualitative and quantitative comparison results of eight different active contour methods (except the
• application specific methods) are presented in [73] for brain MR images, ultrasound pig heart images, kidney CT
• images, knee MR images, and microscopy blood cell images. However, there is no comparative study for liver
• segmentation on SPIR images, which show the vascular structure of the liver very clearly and are very useful for
• vessel segmentation. None of the proposed approaches, namely, the deterministic iterative method (K-means
• based segmentation), the probabilistic model-based iterative method (GMM-based segmentation with EM), the
• supervised learning method (MLP-based segmentation), or the four different level set-based methods have been
• applied to liver segmentation from SPIR images. The contribution of this study is to make a comparison of
• state-of-the-art methods and present them for liver segmentation on abdominal MR images. Seven different
• algorithms have been implemented and the results obtained from SPIR image datasets are presented in this
• section. In addition, we propose an automatic liver segmentation approach using preprocessed images without
• any user interaction or postprocessing operations in Section 6.4.
• The quantitative comparison results given in Table 1 show that the automatic DRLSE with SPF function
• based segmentation approach using preprocessed images presents the most effective segmented liver images with
• the least computational cost among all the applied methods. Efficiency of the regularization of the level set
• function can be increased to get more successful results from this method as a future study. In this way, the
• computational cost of this method can be reduced. Vessel segmentation from the segmented SPIR images will be a future study.
• Caselles V, Kimmel R, Sapiro G. Geodesic active contours. Int Journ Comput Vis 1997; 22: 61–79.
• Malladi R, Sethian JA, Vemuri BC. Shape modeling with front propagation: a level set approach. IEEE Transactions
• on Pattern Analysis and Machine Intelligence 1995; 17: 158–175.
• Vese LA, Chan TF. A multiphase level set framework for image segmentation using the Mumford and Shah model.
• Int Journ Comput Vis 2002; 53: 271–293.
• Wang X, Zheng C, Li C, Yin Y, Feng DD. Automated CT liver segmentation using improved Chan-Vese model
• with global shape constrained energy. 33rd Annual International Conference of the IEEE EMBS 2011; Boston, Massachusetts, USA.
• Zhang Y. Advances in Image and Video Segmentation. Hershey, PA, USA: IRM Press 2006.
• Tsai D. Automatic segmentation of liver structure in CT images using a neural network. IEICE Trans. Fundamentals
• E77–A 1994; 11: 1892–1895.
• Husain SA, Shigeru E. Use of neural networks for feature based recognition of liver region on CT images. In: Neural Networks for Signal Processing - Proceedings of the IEEE Workshop; 2000; 2: pp. 831–840.
• Lee CC, Chung PC, Tsai HM. Identifying multiple abdominal organs from CT image series using a multimodule contextual neural network and spatial fuzzy rules. IEEE Trans Inf Technol Biomedicine 2003; 7: 208–217.
• Bae KT, Giger ML, Chen CT, Kahn CE. Automatic segmentation of liver structure in CT images Med Phys 1993; 20: 71–78.
• Gao L, Heath DG, Kuszyk BS, Fishman EK. Automatic liver segmentation technique for three-dimensional visual
• ization of CT data. Radiology 1996; 201: 359–364.
• Lim SJ, Jeong YY, Ho YS. Automatic liver segmentation for volume measurement in CT images. Journal of Vis.
• Commun Image R 2006; 17: 860–875.
• Chou JS, Chen SY, Sudakoff GS, Hoffmann RK, Chen TC, Dachman AH. Image fusion for visualization of hepatic
• vasculature and tumors. Proc of SPIE 1995; 2434: 157–163.
• Gao J, Kosaka A, Kak A. A deformable model for automatic CT liver extraction. Medical Imaging. SPIE Proc 2000; 2434: pp. 157–163.
• Heimann T, Wolf I, Meinzer HP. Active shape models for a fully automated 3D segmentation of the liver: An
• evaluation on clinical data. In: Proceedings of MICCAI 2006; Lecture Notes in Computer Science 2006; 4191: 41–48.
• Heimann T, Meinzer HP, Wolf I. A statistical deformable model for the segmentation of liver CT volumes. MICCAI 2007. In: Workshop Proceedings: 3D Segmentation in the Clinic: A Grand Challenge 2007; pp. 161–166.
• Pohle R, Toennies KD. Segmentation of medical images using adaptative region growing. Proc SPIE Medical Imaging 2001; 2: 1337–1346.
• Rusk´o L, Bekes G, N´emeth G, Fridrich M. Fully automatic liver segmentation for contrast-enhanced CT images. MICCAI 2007. In: Workshop Proceedings: 3D Segmentation in the Clinic: A Grand Challenge 2007; pp. 143–150.
• Bekes G, Ny¨ul LG, M´at´e E, Kuba A, Fidrich M. 3D segmentation of liver, kidneys and spleen from CT images. International Journal of Computer Assisted Radiology and Surgery 2007; 2: 45–46.
• Shitong W, Duan F, Min X, Dewen H. Advanced fuzzy cellular neural network: Application to CT liver images. Artificial Intelligence in Medicine 2007; 39: 65–77. Suzuki K, Kohlbrenner R, Epstein ML, Obajuluwa AM, Xu J, Hori M. Computer-aided measurement of liver volumes in CT by means of geodesic active contour segmentation coupled with level-set algorithms. Med Phys 2010; 37: 2159–2166.
• Linguraru MG, Sandberg JK, Li Z, Shah F, Summers RM. Automated segmentation and quantification of liver and
• Positano V, Salani B, Scattini B, Santarelli MF, Ramazzotti A, Pepe A, Lombardi M, Landini L. A robust method for assessment of iron overload in liver by magnetic resonance imaging. IEEE Eng. in Medicine and Biology Conf 2007; pp. 2895–2898.
• Lu D, Zhang J, Wang X, Fang J. A fast and robust approach to liver nodule detection in MR images. IEEE Bioscience and Info Tech Conf 2007; pp. 493–497.
• Massoptier L, Casciaro S. Fully automatic liver segmentation through graph-cut technique. IEEE EMBS 29th International Conf 2007; pp. 5243–5246.
• Krishnamurthy C, Rodriguez JJ, Gillies RJ. Snake-based liver lesion segmentation. Conf. Southwest 04 2004. pp.
• Hermoye L, Laamari-Azjal L, Cao Z, Annet L, Lerut J, Dawant BM, Van BBE. Liver segmentation in living liver transplant donors: Comparison of semiautomatic and manual methods. International Journal of Radiology 2005; 234: 171–178.
• Cheng K, Lixu G, Jianghua W, Wei L, Jianrong X. A novel level set based shape prior method for liver segmentation from MRI images. MIAR LNCS 5128 2008. pp. 150–159.
• Strzelecki M, Certaines J, Ko S. Segmentation of 3D MR liver images using synchronised oscillators network. IEEE Info Tech Conf 2007. pp. 259-263.
• Platero C, Gonzalez M, Tobar MC, Poncela JM, Sanguino J, Asensio G, Santas E. Automatic method to segment the liver on multi-phase MRI. Computer Assisted Radiology and Surgery (CARS) 22nd International Congress and Exhibition 2008.
• Fenchel M, Thesen S, Schilling A. Reconstructing liver shape and position from MR image slices using an active shape model. Proc of SPIE 2008.
• Rafiee A, Masoumi H, Roosta A. Using neural network for liver detection in abdominal MRI images. IEEE Int. Conf. on Signal and Image Proc 2009. pp. 21–25.
• Dongxiang C, Tiankun L. Iterative quadtree decomposition segmentation of liver MR image. International Confer- ence on Artificial Intelligence and Computational Intelligence (AICI) 2009; 3: 527–529.
• Tang S, Wang Y. MR-guided liver cancer surgery by nonrigid registration. IEEE International Conf. on Medical Image Analysis and Clinical Application (MIACA) 2010. pp.113–117. Bagci U, Chen X, Udupa JK. Hierarchical scale-based multi-object recognition of 3D anatomical structures. IEEE Transactions on Medical Imaging 2012; 31: 777–789.
• Chen X, Udupa JK, Bagci U, Zhuge Y, Yao J. Medical image segmentation by combining graph cut and oriented
• active appearance models. IEEE Transactions on Image Processing 2012; 21: 2035–2046. Chen X, Bagci U. 3D automatic anatomy segmentation based on iterative graph cut ASM. Medical Physics 2011; 38: 4610–4622.
• Yiming X, Shak M, Francis S, Arnold DL, Arbel T, Collins DL. Optimal Gaussian mixture models of tissue intensities in brain MRI of patients with multiple sclerosis. Machine Learning in Medical Imaging LNCS 2010; 6357: 165–173.
• Chegini M, Ghassemian H. Spatial spectral Gaussian mixture model approach for automatic segmentation of multispectral MR brain images. In: Proceedings of the 19th Iranian Conference on Electrical Engineering (ICEE) 2011. pp. 1–6.
• Dempster AP, Laird N, Rubin D. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical 1977; 39: 1–38. Vincent L. Morphological grayscale reconstruction in image analysis: Applications and efficient algorithms. IEEE Trans Image Process 1993; 2: 176–201.
• Anderberg MR. Cluster Analysis for Applications. New York, NY, USA: Academic Press, 1973.
• Yijun H, Guirong W. Segmentation of cDNA microarray spots using k-means clustering algorithm and mathematical morphology. WASE International Conference on Information Engineering (ICIE 2009) 2009; 2: 110–113.
• Abdul Nazeer KA, Sebastian MP. Improving the accuracy and efficiency of the k-means clustering algorithm. Inter- national Conference on Data Mining and Knowledge Engineering (ICDMKE) Proceedings of the World Congress on Engineering (WCE-2009) 2009; 1: 308–312.
• Juang LH, Wu MN. MRI brain lesion image detection based on color-converted k-means clustering segmentation.
• Measurement 2010; 43: 941–949.
• Yang H, Zhou GT, Yin Y, Yang X. K-means based fingerprint segmentation with sensor interoperability sensor interoperability. EURASIP Journal on Advances in Signal Processing, Article ID: 729378 2010.
• Duda RO, Hart PE. Pattern Classification and Scene Analysis. New York, NY, USA: John Wiley and Sons Inc., 1973.
• MacQueen J. Some methods for classification and analysis of multivariate observations. Proceedings of the Fifth Berkeley Symposium on Mathematics Statistics and Probability I 1967. pp. 281–297.
• Haykin S. A Comprehensive Foundation. Neural Networks. Upper Saddle River, NJ, USA: Prentice-Hall, 1997. pp. 176–201.
• G¨o¸ceri E. Automatic kidney segmentation using Gaussian mixture model on MRI sequences. Electrical Power Systems and Computers 2011; 99: 23–29.
• Otsu N. A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Man and Cybernetics 1979; 9: 62–66. Osher S, Sethian JA. Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulation. J Comput Phys 1988; 79: 12–49.
• Li C, Xu C, Gui C, Fox MD. Distance regularized level set evolution and its application to image segmentation. IEEE Trans Imag Proc 2010; 19: 3243–3254.
• Chan T, Vese L. Active contours without edges. IEEE Trans Image Process 2001; 10: 266–277.
• Caselles V, Catte F, Coll T, Dibos F. A geometric model for active contours in image processing. Numerische
• Mathematik 1993; 66: 1–31.
• Chen G, Gu L, Qian L, Xu J. An improved level set for liver segmentation and perfusion analysis in MRIs. IEEE
• Transactions On Information Technology In Biomedicine 2009; 13: 94–103.
• An J, Rousson M, Xu C. -convergence approximation to piecewise smooth medical image segmentation. Proc MICCAI 2007. pp. 495–502.
• Lankton S, Melonakos J, Malcolm J, Dambreville S, Tannenbaum A. Localized statistics for DW-MRI fiber bundle segmentation. Computer Vision and Pattern Recognition Workshops 2008. pp. 1–8.
• Piovano J, Rousson M, Papadopoulo T. Efficient segmentation of piecewise smooth images. SSVM 2007. pp. 709–720.
• He L, Peng Z, Everding B, Wang X, Han CY, Weiss KL, Wee WG. A comparative study of deformable contour
• methods on medical image segmentation. Image and Vision Computing 2008; 26: 141–163.
• Kimia BB, Tannenbaum A, Zucker SW. Toward a computational theory of shape: An overview. Computer Vision–
• ECCV In: Faugeras O, editor. Lecture Notes in Computer Science 1990; 427: 402–407.
• Rousson M, Paragios N. Shape priors for level set representations. Proceedings of the Seventh European Conference
• on Computer Vision (ECCV) 2002. pp. 78–92.
• Paragios N, Deriche R. Geodesic active region and level set methods for supervised texture segmentation. Int J
• Comput Vis 2002; 46: 223–247.
• Cohen LD, Kimmel R. Global minimum for active contour models: A minimal path approach. Int J Comput Vis 1997; 24: 57–78.
• Chen Y, Thiruvenkadam S, Huang F, Tagare HD, Wilson D, Geiser EA. On the incorporation of shape priors into geometric active contours. Proceedings of the IEEE Workshop on Variational and Level Set Methods 2001. pp.
• Leventon M, Grimson E, Faugeras O. Statistical shape influence in geodesic active contour. Proc IEEE Comput Vis Pattern Recognit (CVPR) 2000. pp. 316–322.
• Wang X, He L, Wee WG. Deformable contour method: a constrained optimization approach. Int J Comput Vis 2004; 59: 87–108.
• Li C, Huang R, Ding, Gatenby JC, Metaxas ND, Gore JC. A level set method for image segmentation in the presence of intensity inhomogeneities with application to MRI. IEEE Transactions on Image Processing 2011; 20: 2007–2016.
• Osher S, Fedkiw R. Level Set Methods and Dynamic Implicit Surfaces. New York, NY, USA: Springer Verlag, 2002.
• Peng D, Merriman B, Osher S, Zhao H, Kang M. A PDE-based fast local level set method. J Comput Phys 1999; 155: 410–438.
• Goldenberg R, Kimmel R, Rivlin E, Rudzsky M. Fast geodesic active contours. IEEE Trans Image Process 2001; 10: 1467–1475.
• Shi Y, Karl WC. A real-time algorithm for the approximation of level-set based curve evolution. IEEE Trans Image Process 2008; 17: 645–656.
• Weeratunga S. and C. Kamath. An investigation of implicit active contours for scientific image segmentation. Video Communications and Image Processing, SPIE; 5308: 210–221.
• G¨o¸ceri E, ¨Unl¨u MZ, G¨uzeli¸s C, Dicle O. An automatic level set based liver segmentation from MRI data sets. In: IPTA 3rd International Conference on Image Processing Theory, Tools and Applications; 15–18 October 2012; ˙Istanbul, Turkey. pp.192–197.