医学图像分割算法研究及其在骨分割中的应用
|
摘要
医学图像分割是医学图像处理到分析过程的关键步骤,在临床医学中发挥着越来越重要的作用。其中,骨分割是医学图像分割在临床医学的主要应用,可用于体积测量、骨骼损伤治疗和修复计划的制定、三维重建等。准确的分割结果可为医学研究和临床诊断提供可靠的数据,以提高医生诊断的准确性。
本文首先分析了医学图像分割技术的发展现状,主要讨论三大类常用的分割算法,包括:基于变形模型的分割方法、基于区域的分割方法和基于统计学的分割方法以及它们的优缺点。体数据分割是基于医学图像分割技术的新方法,其基本思想是:将序列图像构成的体数据看作一个整体进行处理,提高分割的效率和准确率,该方法对三维体视化有着关键作用。针对CT图像骨骼灰度值变化范围大、存在噪声和弱边缘效应等特点给骨分割带来了很大的困难,本文重点对骨分割算法作进一步的研究。
通过算法比较,本文选择分水岭算法对脑部CT图像进行骨分割。从模拟“浸没”过程和集水盆合并两方面进行算法改进以克服过分割现象,从而提高图像的分割质量。对血管造影断层构成的三维体数据,本文提出了一种基于直方图熵的体数据分类方法,通过设置阻光度传递函数的分段点完成去骨目的。
Medical image segmentation is the key step from medical image process to analysis. As the main application in clinical medicine, bone segmentation can be used for measurement of tissue volume, make bone treatment and rehabilitation programs and reconstruct 3D image. Accurate results of segmentation can provide reliable information for medical research and clinical diagnostic.
This paper analyzed the development status of medical image segmentation technology, primarily discussed three commonly algorithms and their advantages and disadvantages. Such as the deformable model-based method, the region-based method, and the statistical_based method. Using 2D image sequences to generate the volume data can improve higher efficiency and accuracy of segmentation results,played an important role in the process of 3D Visualization. There are several characteristics for CT images : variation range of gray values, noise and the weak edge effects, which bring some difficulties for bone segmentation. It is necessary for us to do some further research on bone segmentation.
Compared with segmentation algorithms, this thesis opted the watershed method on brain CT images. To address the problem of over_cut phenomenon, the paper modified the traditional watershed method, including the identification method and the basins merging method in the simulation process of“drown”. In addition, this thesis presented a novel method based on histogram entropy in the volum data classification, adjusting the opacity transfer function to complete the boneless of CTA dates. At last, Using Ray_Casting 3D Visualization technology to display the reconstruct result.
本文首先分析了医学图像分割技术的发展现状,主要讨论三大类常用的分割算法,包括:基于变形模型的分割方法、基于区域的分割方法和基于统计学的分割方法以及它们的优缺点。体数据分割是基于医学图像分割技术的新方法,其基本思想是:将序列图像构成的体数据看作一个整体进行处理,提高分割的效率和准确率,该方法对三维体视化有着关键作用。针对CT图像骨骼灰度值变化范围大、存在噪声和弱边缘效应等特点给骨分割带来了很大的困难,本文重点对骨分割算法作进一步的研究。
通过算法比较,本文选择分水岭算法对脑部CT图像进行骨分割。从模拟“浸没”过程和集水盆合并两方面进行算法改进以克服过分割现象,从而提高图像的分割质量。对血管造影断层构成的三维体数据,本文提出了一种基于直方图熵的体数据分类方法,通过设置阻光度传递函数的分段点完成去骨目的。
Medical image segmentation is the key step from medical image process to analysis. As the main application in clinical medicine, bone segmentation can be used for measurement of tissue volume, make bone treatment and rehabilitation programs and reconstruct 3D image. Accurate results of segmentation can provide reliable information for medical research and clinical diagnostic.
This paper analyzed the development status of medical image segmentation technology, primarily discussed three commonly algorithms and their advantages and disadvantages. Such as the deformable model-based method, the region-based method, and the statistical_based method. Using 2D image sequences to generate the volume data can improve higher efficiency and accuracy of segmentation results,played an important role in the process of 3D Visualization. There are several characteristics for CT images : variation range of gray values, noise and the weak edge effects, which bring some difficulties for bone segmentation. It is necessary for us to do some further research on bone segmentation.
Compared with segmentation algorithms, this thesis opted the watershed method on brain CT images. To address the problem of over_cut phenomenon, the paper modified the traditional watershed method, including the identification method and the basins merging method in the simulation process of“drown”. In addition, this thesis presented a novel method based on histogram entropy in the volum data classification, adjusting the opacity transfer function to complete the boneless of CTA dates. At last, Using Ray_Casting 3D Visualization technology to display the reconstruct result.
引文
[1] Amit Chakraboriy, Lawrenca H.Staib, James S . Deformable Boundary Finding in Medical Imaging by Integrating Gradient and Regin Information. IEEE Trans[J]. Medical Image, 1996, 30(17):859-870.
[2]于志强.医学图像分割与虚拟手术几个关键问题的研究[M].上海:上海交通大学博士学位论文. 2007.
[3] Chao Tian, Mark M, Hod L . Physical Sketching: Reconstruction and analysis of 3D objects from freehand sketches[J]. Computer_Aided Design, 2009, 41(3):147-158.
[4] Salah A, Otman B . Image compression using plane fitting with inter-block prediction[J]. Image and Vision Computing, 2009, 27(4):385-390.
[5] M. Kass, A.Withkin, D. Terzopoulos . Snakes: Active Contour Models. In Proc. 1st Int. Conference on Computer Vision,London, 1987, 22(9):259-268.
[6] Chenyang Xu, Jerry L. Prince . Snakes, Shapes, and Gradient Vector Flow[J]. IEEE Transactions on Imaging Processing, 1998, 7(3):359-369.
[7]田莹,苑玮琦.遗传算法在图像处理中的应用[J].中国图像图形学报, 2007, 12(3):389-392.
[8] Lao Li, Wu Xiaoming, Zhu Xuefeng . Survey on Application of Fuzzy Set Theory for Imaging Segmentation[J]. Chinese Journal of Stereology and Image Analysis, 2006, 11(3):200-205.
[9] J. C. Ma . Segmentation of Multidimension MR Images Using a Fuzzy Neural Network[J]. Processings of SPIE the International Society for Optical Engineering, 1994, 2298(23):636-643.
[10]王成,张海戈,江旭峰,章鲁.基于知识的三维核医学左心室心肌的提取[J].中国生物医学工程报, 2007, 23(4):36-39.
[11] Tina Kapur, Ph.D. Thesis . Model based three dimensional Medical Image Segmentation[J]. Artificial Intelligence Laboratory,Massachusetts Institute of Technonlgy, 1999, 3(10):53-57.
[12] National Electrical Manufacturers Association(NEMA), Digital Imaging and Communications in Medical(DICOM), 2000.
[13]全海英,杨源,张歆东等,DICOM数据集与DCM文件格式[J].计算机应用,2001,21(8): 145-146.
[14]叶峰.医学图像体数据分割及其可视化的研究[M].苏州:硕士学位论文.2008.
[15]林瑶,田捷.医学图像分割方法综述[J].模式识别与人工智能, 2002, 15(2):192-204.
[16] Osher, S., Sethian, J. Fronts Propagating with Curvature Dependent Speed[J]. Algorithms Based on Hamilton-Jacobi Formulation of Comput. Phys, 1997, 79(15):12-49.
[17] Jayaram K. Udupa, Supun Samarasekera . Fuzzy Connectedness and Object Definition: Theory, Algorithms, and Applications in Image Segmentation[J]. Graphical Models and Image Processing, 1996, 58(3). pp:246-261.
[18] Remean A, Borel N . A Region Growing and Merging Algorithm to Color Segmentation[J]. Pattern Recognition, 1994, (23)16:641-647.
[19] Pham D L, Xu C, Prince J L . A Survey of Current Methods in Medical Image Segmentation[J]. Annual Review of Biomedical Engineering, 2000, 12(38):315-338.
[20] Hall L. O, Bensaid A. M, Clarke L.P . A Comparison of Neural Network and Fuzzy Clustering Techniques in Segmenting Magnetic Resonance Images of the Brain[J]. IEEE Trans. Neural Networks, 1992, 30(5):672-681.
[21] BTM Roerdink, J. Meijster, A . Segmentation by Watersheds: Definition and Parallel Implementation[J]. In Advances in Computer Vision, 1997, 3(2):21-30.
[22] Held K, Kops E R, Krause B J, et al . Markov Random Field Segmentation of Brain MR Images[J]. IEEE Transactions on Medical Imaging, 1997, 46(35):878-886.
[23] Michael W. Hansen, William E. Higgins . Relaxation Methods for Supervised Image Segmentation[J]. IEEE Trans Pattern Analysis and Machine Intelligence, 1997, 19(9):949-962.
[24] Gupta, L. and T. Sortrakul . A Gaussian-mixture-based image segmentation algorithm[J]. Pattern Recognition, 1998, 31(11):315-325.
[25]谢杰成,张大力,徐文立.小波图像去噪综述[J].中国图像图形学报, 2002, 7(13):209-217.
[26] Paragios N . Variational methods International Symposium on Biomedical Imaging[J]. From Nano to Micro,Arlington, 2004, 13(5):17-20.
[27] A. P. Witkin . Scale-space filtering[A]. In Proceeding of 8th International Joint Conference on Artificial Intelligence[C], 1983:1019-1022.
[28]王大凯,侯榆青,彭进业.图像处理的偏微分方程[M].北京:科学出版社. 2008.
[29] L. I. Rudin, S. Osher, E. Fatemi . Nonlinear Total Variation Based Noise Removal Algorithms[J]. Physica D, 1992, 60(14):259-268.
[30] F. Zhang, Y.M.Yoo, L. M. Koh, et al . Nonlinear Diffusion in Laplacian Pyramid Domain for Ultrasonic Speckle Reduction[J]. IEEE Transactions on Medical Imaging, 2007, 26(2):200-211.
[32] S Beucher, C Lantuejoul . Use of Watersheds in Contour Detection[J]. International Workshop on Image Processing: Real-time Edge and Motion Detection/Estimation, 1979, 11(9):17-21.
[32]韩峰.一种自适应分水岭数字图像分割技术研究[M].长沙:湖南大学硕士学位论文, 2007.
[33] L Vincent, P Soille . Watersheds in Digital Space: An Efficient Algorithm dased on Immersion Simulation[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 1991, 13(6):583-598.
[34]刘晨,张东.边缘检测算子研究及其在医学图像中的应用[J].计算机技术及发展, 2006, 16(8):128-131.
[35]崔屹.图像处理与分析:数学形态学方法及应用[A].北京:科学出版社, 2000.
[36]王宇,陈殿仁,沈美丽等.基于形态学梯度重构和标记提取的分水岭图像分割[J] .中国图象图形学报, 2008, 13(11):-2180.
[37] P Soille . Morphological Image Analysis Principles and Applications[A]. Berlin, Germany: Springer Verlag, 2003.
[38]邓子健,李弼程.基于直观分水岭定义的图像分割算法[J].计算机工程与应用, 2005, 41(26):43-47.
[39] H K Hahn, H O Peitgen . IWT-Interactive Watershed Transform: A hierarchical method for efficient interactive and automated segmentation of multidimensional grayscale images[J]. Proc on SPIE Medical Imaging, 2003, 50(32):643-653.
[40]高丽,杨树元,李海强.一种基于标记的分水岭图像分割新算法[J].中国图象图形学报, 2007, 6(12):1025-1032.
[41] Parveen, Runa, Todd-Pokropek, et al. Classification of MR Brain Tissues Using Fuzzy Estimation[J]. Nuclear Science Symposium Conference Record, 2006, 13(4):2613-2619.
[42] Shin B S . An Efficient Classification and Rendering Method Using Tagged Distance Maps[J]. Visual Computer, 2004, 20(8):540-553.
[43] Lin Xue-yan, Liu Zheng-guang . Classification and Rendering of Brain MRImaging Based on Wavelet-domain Hidden Markov Mode[C]. Proc of International Conference on Imaging: Technology and Applications for 21st Century, 2005:342-343.
[44] Pohle R, Bischof L . Seeded Region Growing[J]. IEEE Transcations on Pattern Analysis and Machine Intelligence, 1994, 16(6):641-647.
[45] Pohle R, Bischof L . A New Approach for Model-based Adaptive Region Growing in Medical Image Analysis[A]. Proceedings of the 19th International Conference on Computer Analysis and Patterns, 2001:238-246.
[46]陆剑锋,林海,潘志庚.自适应区域生长算法在医学图像分割中的应用[J].计算机辅助设计与图形学学报, 2005, 17(10):2168-2173.
[47] Evoy M . Volume Rendering: Display of Surfaces from Volume Data[J]. IEEE Computer Graphics and Applications, 1988, 8(5):29-37.
[48]杨静宇,曹雨龙.计算机图像处理及常用算法手册[K].南京:南京大学出版社, 1997:173-209.
[49] Natwong B, Sooraksa P, Pintavriooj C, et al . Wavelet Entropy Analysis of the Resolution ECG[C]. Proc of the 1st IEEE Conference on Industrial Electronics and Applications, 2006:1-4.
[50] Kapur J N, Sahoo P K, Wong A K C . A New Method for Gray level Picture Thresholding using the Entropy of the Histogram[J]. Computer Vision,Graphics, and Image Processing, 1985, 29(3):273-285.
[2]于志强.医学图像分割与虚拟手术几个关键问题的研究[M].上海:上海交通大学博士学位论文. 2007.
[3] Chao Tian, Mark M, Hod L . Physical Sketching: Reconstruction and analysis of 3D objects from freehand sketches[J]. Computer_Aided Design, 2009, 41(3):147-158.
[4] Salah A, Otman B . Image compression using plane fitting with inter-block prediction[J]. Image and Vision Computing, 2009, 27(4):385-390.
[5] M. Kass, A.Withkin, D. Terzopoulos . Snakes: Active Contour Models. In Proc. 1st Int. Conference on Computer Vision,London, 1987, 22(9):259-268.
[6] Chenyang Xu, Jerry L. Prince . Snakes, Shapes, and Gradient Vector Flow[J]. IEEE Transactions on Imaging Processing, 1998, 7(3):359-369.
[7]田莹,苑玮琦.遗传算法在图像处理中的应用[J].中国图像图形学报, 2007, 12(3):389-392.
[8] Lao Li, Wu Xiaoming, Zhu Xuefeng . Survey on Application of Fuzzy Set Theory for Imaging Segmentation[J]. Chinese Journal of Stereology and Image Analysis, 2006, 11(3):200-205.
[9] J. C. Ma . Segmentation of Multidimension MR Images Using a Fuzzy Neural Network[J]. Processings of SPIE the International Society for Optical Engineering, 1994, 2298(23):636-643.
[10]王成,张海戈,江旭峰,章鲁.基于知识的三维核医学左心室心肌的提取[J].中国生物医学工程报, 2007, 23(4):36-39.
[11] Tina Kapur, Ph.D. Thesis . Model based three dimensional Medical Image Segmentation[J]. Artificial Intelligence Laboratory,Massachusetts Institute of Technonlgy, 1999, 3(10):53-57.
[12] National Electrical Manufacturers Association(NEMA), Digital Imaging and Communications in Medical(DICOM), 2000.
[13]全海英,杨源,张歆东等,DICOM数据集与DCM文件格式[J].计算机应用,2001,21(8): 145-146.
[14]叶峰.医学图像体数据分割及其可视化的研究[M].苏州:硕士学位论文.2008.
[15]林瑶,田捷.医学图像分割方法综述[J].模式识别与人工智能, 2002, 15(2):192-204.
[16] Osher, S., Sethian, J. Fronts Propagating with Curvature Dependent Speed[J]. Algorithms Based on Hamilton-Jacobi Formulation of Comput. Phys, 1997, 79(15):12-49.
[17] Jayaram K. Udupa, Supun Samarasekera . Fuzzy Connectedness and Object Definition: Theory, Algorithms, and Applications in Image Segmentation[J]. Graphical Models and Image Processing, 1996, 58(3). pp:246-261.
[18] Remean A, Borel N . A Region Growing and Merging Algorithm to Color Segmentation[J]. Pattern Recognition, 1994, (23)16:641-647.
[19] Pham D L, Xu C, Prince J L . A Survey of Current Methods in Medical Image Segmentation[J]. Annual Review of Biomedical Engineering, 2000, 12(38):315-338.
[20] Hall L. O, Bensaid A. M, Clarke L.P . A Comparison of Neural Network and Fuzzy Clustering Techniques in Segmenting Magnetic Resonance Images of the Brain[J]. IEEE Trans. Neural Networks, 1992, 30(5):672-681.
[21] BTM Roerdink, J. Meijster, A . Segmentation by Watersheds: Definition and Parallel Implementation[J]. In Advances in Computer Vision, 1997, 3(2):21-30.
[22] Held K, Kops E R, Krause B J, et al . Markov Random Field Segmentation of Brain MR Images[J]. IEEE Transactions on Medical Imaging, 1997, 46(35):878-886.
[23] Michael W. Hansen, William E. Higgins . Relaxation Methods for Supervised Image Segmentation[J]. IEEE Trans Pattern Analysis and Machine Intelligence, 1997, 19(9):949-962.
[24] Gupta, L. and T. Sortrakul . A Gaussian-mixture-based image segmentation algorithm[J]. Pattern Recognition, 1998, 31(11):315-325.
[25]谢杰成,张大力,徐文立.小波图像去噪综述[J].中国图像图形学报, 2002, 7(13):209-217.
[26] Paragios N . Variational methods International Symposium on Biomedical Imaging[J]. From Nano to Micro,Arlington, 2004, 13(5):17-20.
[27] A. P. Witkin . Scale-space filtering[A]. In Proceeding of 8th International Joint Conference on Artificial Intelligence[C], 1983:1019-1022.
[28]王大凯,侯榆青,彭进业.图像处理的偏微分方程[M].北京:科学出版社. 2008.
[29] L. I. Rudin, S. Osher, E. Fatemi . Nonlinear Total Variation Based Noise Removal Algorithms[J]. Physica D, 1992, 60(14):259-268.
[30] F. Zhang, Y.M.Yoo, L. M. Koh, et al . Nonlinear Diffusion in Laplacian Pyramid Domain for Ultrasonic Speckle Reduction[J]. IEEE Transactions on Medical Imaging, 2007, 26(2):200-211.
[32] S Beucher, C Lantuejoul . Use of Watersheds in Contour Detection[J]. International Workshop on Image Processing: Real-time Edge and Motion Detection/Estimation, 1979, 11(9):17-21.
[32]韩峰.一种自适应分水岭数字图像分割技术研究[M].长沙:湖南大学硕士学位论文, 2007.
[33] L Vincent, P Soille . Watersheds in Digital Space: An Efficient Algorithm dased on Immersion Simulation[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 1991, 13(6):583-598.
[34]刘晨,张东.边缘检测算子研究及其在医学图像中的应用[J].计算机技术及发展, 2006, 16(8):128-131.
[35]崔屹.图像处理与分析:数学形态学方法及应用[A].北京:科学出版社, 2000.
[36]王宇,陈殿仁,沈美丽等.基于形态学梯度重构和标记提取的分水岭图像分割[J] .中国图象图形学报, 2008, 13(11):-2180.
[37] P Soille . Morphological Image Analysis Principles and Applications[A]. Berlin, Germany: Springer Verlag, 2003.
[38]邓子健,李弼程.基于直观分水岭定义的图像分割算法[J].计算机工程与应用, 2005, 41(26):43-47.
[39] H K Hahn, H O Peitgen . IWT-Interactive Watershed Transform: A hierarchical method for efficient interactive and automated segmentation of multidimensional grayscale images[J]. Proc on SPIE Medical Imaging, 2003, 50(32):643-653.
[40]高丽,杨树元,李海强.一种基于标记的分水岭图像分割新算法[J].中国图象图形学报, 2007, 6(12):1025-1032.
[41] Parveen, Runa, Todd-Pokropek, et al. Classification of MR Brain Tissues Using Fuzzy Estimation[J]. Nuclear Science Symposium Conference Record, 2006, 13(4):2613-2619.
[42] Shin B S . An Efficient Classification and Rendering Method Using Tagged Distance Maps[J]. Visual Computer, 2004, 20(8):540-553.
[43] Lin Xue-yan, Liu Zheng-guang . Classification and Rendering of Brain MRImaging Based on Wavelet-domain Hidden Markov Mode[C]. Proc of International Conference on Imaging: Technology and Applications for 21st Century, 2005:342-343.
[44] Pohle R, Bischof L . Seeded Region Growing[J]. IEEE Transcations on Pattern Analysis and Machine Intelligence, 1994, 16(6):641-647.
[45] Pohle R, Bischof L . A New Approach for Model-based Adaptive Region Growing in Medical Image Analysis[A]. Proceedings of the 19th International Conference on Computer Analysis and Patterns, 2001:238-246.
[46]陆剑锋,林海,潘志庚.自适应区域生长算法在医学图像分割中的应用[J].计算机辅助设计与图形学学报, 2005, 17(10):2168-2173.
[47] Evoy M . Volume Rendering: Display of Surfaces from Volume Data[J]. IEEE Computer Graphics and Applications, 1988, 8(5):29-37.
[48]杨静宇,曹雨龙.计算机图像处理及常用算法手册[K].南京:南京大学出版社, 1997:173-209.
[49] Natwong B, Sooraksa P, Pintavriooj C, et al . Wavelet Entropy Analysis of the Resolution ECG[C]. Proc of the 1st IEEE Conference on Industrial Electronics and Applications, 2006:1-4.
[50] Kapur J N, Sahoo P K, Wong A K C . A New Method for Gray level Picture Thresholding using the Entropy of the Histogram[J]. Computer Vision,Graphics, and Image Processing, 1985, 29(3):273-285.