基于CT图像肺部异常阴影的自动检测
|
摘要
恶性肿瘤是当今人类面临的一个非常严峻的健康问题,而肺癌是所有恶性肿瘤中对人类危害最大的。早期发现肺部肿瘤对肺癌的治疗意义重大,目前临床上主要采用X射线CT机检测早期的肺部病变,然而随着CT设备的不断改进,病人的CT摄片数越来越多,这不仅加大了医生的工作负担,也增加了漏诊和误诊发生的几率。本文研究了基于CT图像肺部异常阴影的自动检测技术,这也是肺部CAD系统中最关键的一部分。
胸部CT图像中肺实质的分割是异常阴影检测的前提。本文首先通过边界跟踪算法与洪水填充算法提取出了CT图像中的胸部区域,进而从二值化后的胸部区域中找到初始肺部边界。为了保证分割出的肺部区域的完整性,本文提出了一种新的基于弧长表示的曲线平滑算法,用于平滑初始肺部边界,然后给出了一种新的边界修补算法,用于修补平滑后的肺部边界,最终实现肺部区域的完整分割。
本文主要采用基于数学形态学膨胀运算的环形滤波器进行异常阴影的检测。文中首先介绍了二维环形滤波器的原理,为了过滤掉那些由血管造成的假阳性结果,本文将环形滤波器推广到了三维,然后针对环形滤波器的不足之处,提出了三种改进措施,最终得到了适应性更强的可变环形滤波器。本文对改进过程中遇到的图像距离变换问题还进行了深入的研究,给出了一种高效的实现二值图像欧氏距离变换的算法,同时还定义了一种新的路径长度,用以实现灰度图像的加权距离变换。实验结果表明,该检测系统可以大大减轻医生的工作负担。
Malignant tumor is a very serious health problem for human beings and lung canceris the most harmful one at present. Lung nodule detection in early stage is very importantfor lung cancer treatment. And X-ray computed tomography (CT) is the most sensitivemodality for lung nodule detection in clinical. However, with the continuous improvementof CT equipment, it produces more and more images. Large amount of CT images notonly burden the image interpretation physician, but also increase the rate of overlook andmisdiagnosis. This dissertation focuses on computer-aided detection of lung nodules inchest CT images which is the most important part of lung CAD system.
Segmentation of lungs from CT images is the precondition of lung nodules detection.At first, thorax region is segmented from the binarized CT images through border tracingalgorithm and ?ood fill algorithm. Then initial lung border is found in the binarized thoraxregion. In order to ensure the integrity of lung region, a new smoothing algorithm whichis based on the curve expression that take arc-length as a parameter is presented and it isused to smooth the initial lung border. Another lung border correction algorithm which isused to correct the smoothed lung border has also been proposed. At the end, lung regionis segmented completely.
In this dissertation quoit filter which is based on Mathematical Morphology dilationis used to detect lung nodules. At the beginning, the principle of two-dimensional quoitfilter is introduced and then quoit filter is extended to three-dimensional in order to getrid of those false positive results caused by the vessels. For the inadequacies of quoitfilter, three improvements are proposed and ultimately obtained more adaptive variablequoit filter. The problem of distance transform of an image which is encountered in theimprovement process has also been studied in this paper. An e?cient Euclidean distancetransform algorithm has been given and a new path length is defined to achieve the gray-level weighted distance transform. Results of the experiment showed that this detectionsystem can drastically reduce the burden of the doctor.
胸部CT图像中肺实质的分割是异常阴影检测的前提。本文首先通过边界跟踪算法与洪水填充算法提取出了CT图像中的胸部区域,进而从二值化后的胸部区域中找到初始肺部边界。为了保证分割出的肺部区域的完整性,本文提出了一种新的基于弧长表示的曲线平滑算法,用于平滑初始肺部边界,然后给出了一种新的边界修补算法,用于修补平滑后的肺部边界,最终实现肺部区域的完整分割。
本文主要采用基于数学形态学膨胀运算的环形滤波器进行异常阴影的检测。文中首先介绍了二维环形滤波器的原理,为了过滤掉那些由血管造成的假阳性结果,本文将环形滤波器推广到了三维,然后针对环形滤波器的不足之处,提出了三种改进措施,最终得到了适应性更强的可变环形滤波器。本文对改进过程中遇到的图像距离变换问题还进行了深入的研究,给出了一种高效的实现二值图像欧氏距离变换的算法,同时还定义了一种新的路径长度,用以实现灰度图像的加权距离变换。实验结果表明,该检测系统可以大大减轻医生的工作负担。
Malignant tumor is a very serious health problem for human beings and lung canceris the most harmful one at present. Lung nodule detection in early stage is very importantfor lung cancer treatment. And X-ray computed tomography (CT) is the most sensitivemodality for lung nodule detection in clinical. However, with the continuous improvementof CT equipment, it produces more and more images. Large amount of CT images notonly burden the image interpretation physician, but also increase the rate of overlook andmisdiagnosis. This dissertation focuses on computer-aided detection of lung nodules inchest CT images which is the most important part of lung CAD system.
Segmentation of lungs from CT images is the precondition of lung nodules detection.At first, thorax region is segmented from the binarized CT images through border tracingalgorithm and ?ood fill algorithm. Then initial lung border is found in the binarized thoraxregion. In order to ensure the integrity of lung region, a new smoothing algorithm whichis based on the curve expression that take arc-length as a parameter is presented and it isused to smooth the initial lung border. Another lung border correction algorithm which isused to correct the smoothed lung border has also been proposed. At the end, lung regionis segmented completely.
In this dissertation quoit filter which is based on Mathematical Morphology dilationis used to detect lung nodules. At the beginning, the principle of two-dimensional quoitfilter is introduced and then quoit filter is extended to three-dimensional in order to getrid of those false positive results caused by the vessels. For the inadequacies of quoitfilter, three improvements are proposed and ultimately obtained more adaptive variablequoit filter. The problem of distance transform of an image which is encountered in theimprovement process has also been studied in this paper. An e?cient Euclidean distancetransform algorithm has been given and a new path length is defined to achieve the gray-level weighted distance transform. Results of the experiment showed that this detectionsystem can drastically reduce the burden of the doctor.
引文
[1]王娟娟.胸部X线影像中肺部结节的自动检测技术-不平衡数据集的分析[D].上海:上海交通大学生命科学技术学院, 2007: 1-2.
[2]杨玲,李连弟,陈育德.中国肺癌死亡趋势分析及发病、死亡的估计与预测[J].中国肺癌杂志, 2005, 8(4): 274-278.
[3] Shields P G. Molecular Epidemiology of Smoking and Lung Cancer[J]. Oncogene,2002, 21(45): 6870-6876.
[4] Bo?etta P, Trichopoulos D. Cancer of the Lung, Larynx, and Pleura[M]//Adami HO, Hunter D, Trichopoulos D. Textbook of Cancer Epidemiology. England: OxfordUniversity Press, 2008, 349-377.
[5] Muhm J R, Brown L R, Crowe J K. Detection of Pulmonary Nodules by ComputedTomography[J]. American Journal of Roentgenology, 1977, 128(2): 267-270.
[6] Naidich D P, Rusinek H, McGuinness G, et al. Variables A?ecting Pulmonary N-odule Detection with Computed Tomography: Evaluation With Three-dimensionalComputer Simulation[J]. Journal of Thoracic Imaging, 1993, 8(4): 291-299.
[7] Gurney J W. Missed Lung Cancer at CT: Imaging Findings in Nine Patients[J]. Ra-diology, 1996, 199(1): 117-122.
[8] Zhou X, Hayashi T, Hara T, et al. Automatic Segmentation and Recognition ofAnatomical Lung Structures from High-resolution Chest CT Images[J]. Comput-erized Medical Imaging and Graphics, 2006, 30(5): 299-313.
[9] da Silva Sousa J R F, Silva A C, De Paiva A C, et al. Methodology for Automat-ic Detection of Lung Nodules in Computerized Tomography Images[J]. ComputerMethods and Programs In Biomedicine, 2010, 98(1): 1-14.
[10] Hong H, Lee J, Yim Y. Automatic Lung Nodule Matching on Sequential CT Im-ages[J]. Computers in Biology and Medicine, 2008, 38(5): 623-634.
[11] Yim Y, Hong H. Correction of Segmented Lung Boundary for Inclusion of PleuralNodules and Pulmonary Vessels in Chest CT Images[J]. Computers in Biology andMedicine, 2008, 38(8): 845-857.
[12] Yim Y, Hong H, Beom Seo J, et al. Correction of Lung Boundary Using the Gradi-ent and Intensity Distribution[J]. Computers in Biology and Medicine, 2009, 39(3):239-250.
[13] Itai Y, Kim H, Ishikawa S, et al. A Segmentation Method of Lung Area by UsingSnakes and Automatic Detection of Abnormal Shadow on the Area[J]. InternationalJournal of Innovative Computing, Information and Control, 2007, 3(2): 277-284.
[14] Bellotti R, De Carlo F, Gargano G, et al. A CAD System for Nodule Detection inLow-dose Lung CTs Based on Region Growing and a New Active Contour Mod-el[J]. Medical Physics, 2007, 34(12): 4901-4910.
[15] Kakar M, Olsen D R. Automatic Segmentation and Recognition of Lungs and Lesionfrom CT Scans of Thorax[J]. Computerized Medical Imaging and Graphics, 2009,33(1): 72-82.
[16] Denison D M, Morgan M D, Millar A B. Estimation of Regional Gas and Tissue Vol-umes of the Lung in Supine Man Using Computed Tomography[J]. Thorax, 1986,41(8): 620-628.
[17] Kalendar W A, Fichte H, Bautz W, et al. Semiautomatic Evaluation Procedures forQuantitative CT of the Lung[J]. Journal of Computer Assisted Tomography, 1991,15(2): 248-255.
[18] Armato S G, Giger M L, Moran C J, et al. Computerized Detection of PulmonaryNodules on CT Scans[J]. Radiographics, 1999, 19(5): 1303-1311.
[19] Bae K T, Kim J S, Na Y H, et al. Pulmonary Nodules: Automated Detection on CTImages with Morphologic Matching Algorithm-Preliminary Results[J]. Radiology,2005, 236(1): 286-294.
[20] Kim D Y, Kim J H, Noh S M, et al. Pulmonary Nodule Detection Using Chest CTImages[J]. Acta Radiologica, 2003, 44(3): 252-257.
[21] Dajnowiec M, Alirezaie J. Computer Simulation for Segmentation of Lung Nodulesin CT Images[C]//IEEE International Conference on Systems, Man and Cybernetics.Hague, 2004, 5: 4491-4496.
[22] Zhang X, McLennan G, Ho?man E A, et al. Automatic Detection of Small-sizePulmonary Nodules Based on Helical CT Images[C]//The 19th International Con-ference on Information Processing in Medical Imaging. Colorado, 2005: 664-676.
[23] Wiemker R, Rogalla P, Zwartkruis A, et al. Computer Aided Lung Nodule Detectionon High Resolution CT Data[C]//SPIE Medical Imaging 2002: Image Processing.San Diego, 2002, 4684(3): 677-688.
[24] Ko J P, Betke M. Chest CT: Automated Nodule Detection and Assessment of Changeover Time: Preliminary Experience[J]. Radiology, 2001, 218(1): 267-273.
[25] Pu J, Roos J, Yi C A, et al. Adaptive Border Marching Algorithm: Automatic LungSegmentation on Chest CT Images[J]. Computerized Medical Imaging and Graph-ics, 2008, 32(6): 452-462.
[26] Bae K T, Giger M, MacMahon H, et al. Computer-aided Detection of PulmonaryNodules in CT Images[J]. Radiology, 1991, 181: 144.
[27] Messay T, Hardie R C, Rogers S K. A New Computationally E?cient CAD Systemfor Pulmonary Nodule Detection in CT Imagery[J]. Medical Image Analysis, 2010,14(3): 390-406.
[28] Dehmeshki J, Ye X, Lin X, et al. Automated Detection of Lung Nodules in CTImages Using Shape-based Genetic Algorithm[J]. Computerized Medical Imagingand Graphics, 2007, 31(6): 408-417.
[29] Lee Y, Hara T, Fujita H, et al. Automated Detection of Pulmonary Nodules in HelicalCT Images Based on an Improved Template-matching Technique[J]. IEEE Transac-tions on Medical Imaging, 2001, 20(7): 595-604.
[30] Fotin S V, Reeves A P, Biancardi A M, et al. A Multiscale Laplacian of GaussianFiltering Approach to Automated Pulmonary Nodule Detection from Whole-lungLow-dose CT Scans[C]//SPIE Medical Imaging 2009: Computer-Aided Diagnosis.Orlando, 2009, 7260: 1-8.
[31] Yamamoto S, Matsumoto M, Tateno Y, et al. Quoit Filter: A New Filter Based onMathematical Morphology to Extract the Isolated Shadow, and Its Application toAutomatic Detection of Lung Cancer in X-ray CT[C]//The 13th International Con-ference on Pattern Recognition. Vienna, 1996, 2: 3-7.
[32] Miwa T, Kako J, Yamamoto S, et al. Automatic Detection of Lung Cancers in ChestCT Images by the Variable N-quoit Filter[J]. Systems and Computers in Japan, 2002,33(1): 53-63.
[33] Yamamoto S. Image Processing Algorithm of Computer-aided Diagnosis in LungCancer Screening by CT[J]. Systems and Computers in Japan, 2005, 36(7): 40-53.
[34] Sua′rez-Cuenca J J, Tahoces P G, Souto M, et al. Application of the Iris Filter forAutomatic Detection of Pulmonary Nodules on Computed Tomography Images[J].Computer in Biology and Medicine, 2009, 39(10): 921-933.
[35] Kubota T, Jerebko A K, Dewan M, et al. Segmentation of Pulmonary Nodules ofVarious Densities with Morphological Approaches and Convexity Models[J]. Med-ical Image Analysis, 2011, 15(1): 133-154.
[36] Pu J, Zheng B, Leader J K, et al. An Automated CT Based Lung Nodule DetectionScheme Using Geometric Analysis of Signed Distance Field[J]. Medical Physics,2008, 35(8): 3453-3461.
[37] Soille P. Morphological Image Analysis: Principles and Applications[M]. Berlin:Springer Press, 2003: 1-7.
[38] Otsu N. A Threshold Selection Method from Gray Level Histograms[J]. IEEE Trans-actions on Systems, Man, and Cybernetics, 1979, 9(1): 62-66.
[39] Sonka M, Hlavac V, Boyle R. Image Processing, Analysis, and Machine Vision[M].New Jersey: Thomson Learning and PT Press, 1999: 96-100.
[40] Rogers D F.原著,石教英等译. Procedural Elements for Computer Graphics,计算机图形学的算法基础[M].北京:机械工业出版社, 2002: 99-106.
[41] Shih F Y, Wu Y T. Fast Euclidean Distance Transformation in Two Scans Usinga 3×3 Neighborhood[J]. Computer Vision and Image Understanding, 2004, 93(2):195-205.
[42] Toivanen P J. New Geodosic Distance Transforms for Gray-scale Images[J]. PatternRecognition Letters, 1996, 17(5): 437-450.
[43] Gonzalez R C, Woods R E. Digital Images Processing Using MATLAB[M]. NewYork: Addison Wesley Press, 2004: 308-310.
[2]杨玲,李连弟,陈育德.中国肺癌死亡趋势分析及发病、死亡的估计与预测[J].中国肺癌杂志, 2005, 8(4): 274-278.
[3] Shields P G. Molecular Epidemiology of Smoking and Lung Cancer[J]. Oncogene,2002, 21(45): 6870-6876.
[4] Bo?etta P, Trichopoulos D. Cancer of the Lung, Larynx, and Pleura[M]//Adami HO, Hunter D, Trichopoulos D. Textbook of Cancer Epidemiology. England: OxfordUniversity Press, 2008, 349-377.
[5] Muhm J R, Brown L R, Crowe J K. Detection of Pulmonary Nodules by ComputedTomography[J]. American Journal of Roentgenology, 1977, 128(2): 267-270.
[6] Naidich D P, Rusinek H, McGuinness G, et al. Variables A?ecting Pulmonary N-odule Detection with Computed Tomography: Evaluation With Three-dimensionalComputer Simulation[J]. Journal of Thoracic Imaging, 1993, 8(4): 291-299.
[7] Gurney J W. Missed Lung Cancer at CT: Imaging Findings in Nine Patients[J]. Ra-diology, 1996, 199(1): 117-122.
[8] Zhou X, Hayashi T, Hara T, et al. Automatic Segmentation and Recognition ofAnatomical Lung Structures from High-resolution Chest CT Images[J]. Comput-erized Medical Imaging and Graphics, 2006, 30(5): 299-313.
[9] da Silva Sousa J R F, Silva A C, De Paiva A C, et al. Methodology for Automat-ic Detection of Lung Nodules in Computerized Tomography Images[J]. ComputerMethods and Programs In Biomedicine, 2010, 98(1): 1-14.
[10] Hong H, Lee J, Yim Y. Automatic Lung Nodule Matching on Sequential CT Im-ages[J]. Computers in Biology and Medicine, 2008, 38(5): 623-634.
[11] Yim Y, Hong H. Correction of Segmented Lung Boundary for Inclusion of PleuralNodules and Pulmonary Vessels in Chest CT Images[J]. Computers in Biology andMedicine, 2008, 38(8): 845-857.
[12] Yim Y, Hong H, Beom Seo J, et al. Correction of Lung Boundary Using the Gradi-ent and Intensity Distribution[J]. Computers in Biology and Medicine, 2009, 39(3):239-250.
[13] Itai Y, Kim H, Ishikawa S, et al. A Segmentation Method of Lung Area by UsingSnakes and Automatic Detection of Abnormal Shadow on the Area[J]. InternationalJournal of Innovative Computing, Information and Control, 2007, 3(2): 277-284.
[14] Bellotti R, De Carlo F, Gargano G, et al. A CAD System for Nodule Detection inLow-dose Lung CTs Based on Region Growing and a New Active Contour Mod-el[J]. Medical Physics, 2007, 34(12): 4901-4910.
[15] Kakar M, Olsen D R. Automatic Segmentation and Recognition of Lungs and Lesionfrom CT Scans of Thorax[J]. Computerized Medical Imaging and Graphics, 2009,33(1): 72-82.
[16] Denison D M, Morgan M D, Millar A B. Estimation of Regional Gas and Tissue Vol-umes of the Lung in Supine Man Using Computed Tomography[J]. Thorax, 1986,41(8): 620-628.
[17] Kalendar W A, Fichte H, Bautz W, et al. Semiautomatic Evaluation Procedures forQuantitative CT of the Lung[J]. Journal of Computer Assisted Tomography, 1991,15(2): 248-255.
[18] Armato S G, Giger M L, Moran C J, et al. Computerized Detection of PulmonaryNodules on CT Scans[J]. Radiographics, 1999, 19(5): 1303-1311.
[19] Bae K T, Kim J S, Na Y H, et al. Pulmonary Nodules: Automated Detection on CTImages with Morphologic Matching Algorithm-Preliminary Results[J]. Radiology,2005, 236(1): 286-294.
[20] Kim D Y, Kim J H, Noh S M, et al. Pulmonary Nodule Detection Using Chest CTImages[J]. Acta Radiologica, 2003, 44(3): 252-257.
[21] Dajnowiec M, Alirezaie J. Computer Simulation for Segmentation of Lung Nodulesin CT Images[C]//IEEE International Conference on Systems, Man and Cybernetics.Hague, 2004, 5: 4491-4496.
[22] Zhang X, McLennan G, Ho?man E A, et al. Automatic Detection of Small-sizePulmonary Nodules Based on Helical CT Images[C]//The 19th International Con-ference on Information Processing in Medical Imaging. Colorado, 2005: 664-676.
[23] Wiemker R, Rogalla P, Zwartkruis A, et al. Computer Aided Lung Nodule Detectionon High Resolution CT Data[C]//SPIE Medical Imaging 2002: Image Processing.San Diego, 2002, 4684(3): 677-688.
[24] Ko J P, Betke M. Chest CT: Automated Nodule Detection and Assessment of Changeover Time: Preliminary Experience[J]. Radiology, 2001, 218(1): 267-273.
[25] Pu J, Roos J, Yi C A, et al. Adaptive Border Marching Algorithm: Automatic LungSegmentation on Chest CT Images[J]. Computerized Medical Imaging and Graph-ics, 2008, 32(6): 452-462.
[26] Bae K T, Giger M, MacMahon H, et al. Computer-aided Detection of PulmonaryNodules in CT Images[J]. Radiology, 1991, 181: 144.
[27] Messay T, Hardie R C, Rogers S K. A New Computationally E?cient CAD Systemfor Pulmonary Nodule Detection in CT Imagery[J]. Medical Image Analysis, 2010,14(3): 390-406.
[28] Dehmeshki J, Ye X, Lin X, et al. Automated Detection of Lung Nodules in CTImages Using Shape-based Genetic Algorithm[J]. Computerized Medical Imagingand Graphics, 2007, 31(6): 408-417.
[29] Lee Y, Hara T, Fujita H, et al. Automated Detection of Pulmonary Nodules in HelicalCT Images Based on an Improved Template-matching Technique[J]. IEEE Transac-tions on Medical Imaging, 2001, 20(7): 595-604.
[30] Fotin S V, Reeves A P, Biancardi A M, et al. A Multiscale Laplacian of GaussianFiltering Approach to Automated Pulmonary Nodule Detection from Whole-lungLow-dose CT Scans[C]//SPIE Medical Imaging 2009: Computer-Aided Diagnosis.Orlando, 2009, 7260: 1-8.
[31] Yamamoto S, Matsumoto M, Tateno Y, et al. Quoit Filter: A New Filter Based onMathematical Morphology to Extract the Isolated Shadow, and Its Application toAutomatic Detection of Lung Cancer in X-ray CT[C]//The 13th International Con-ference on Pattern Recognition. Vienna, 1996, 2: 3-7.
[32] Miwa T, Kako J, Yamamoto S, et al. Automatic Detection of Lung Cancers in ChestCT Images by the Variable N-quoit Filter[J]. Systems and Computers in Japan, 2002,33(1): 53-63.
[33] Yamamoto S. Image Processing Algorithm of Computer-aided Diagnosis in LungCancer Screening by CT[J]. Systems and Computers in Japan, 2005, 36(7): 40-53.
[34] Sua′rez-Cuenca J J, Tahoces P G, Souto M, et al. Application of the Iris Filter forAutomatic Detection of Pulmonary Nodules on Computed Tomography Images[J].Computer in Biology and Medicine, 2009, 39(10): 921-933.
[35] Kubota T, Jerebko A K, Dewan M, et al. Segmentation of Pulmonary Nodules ofVarious Densities with Morphological Approaches and Convexity Models[J]. Med-ical Image Analysis, 2011, 15(1): 133-154.
[36] Pu J, Zheng B, Leader J K, et al. An Automated CT Based Lung Nodule DetectionScheme Using Geometric Analysis of Signed Distance Field[J]. Medical Physics,2008, 35(8): 3453-3461.
[37] Soille P. Morphological Image Analysis: Principles and Applications[M]. Berlin:Springer Press, 2003: 1-7.
[38] Otsu N. A Threshold Selection Method from Gray Level Histograms[J]. IEEE Trans-actions on Systems, Man, and Cybernetics, 1979, 9(1): 62-66.
[39] Sonka M, Hlavac V, Boyle R. Image Processing, Analysis, and Machine Vision[M].New Jersey: Thomson Learning and PT Press, 1999: 96-100.
[40] Rogers D F.原著,石教英等译. Procedural Elements for Computer Graphics,计算机图形学的算法基础[M].北京:机械工业出版社, 2002: 99-106.
[41] Shih F Y, Wu Y T. Fast Euclidean Distance Transformation in Two Scans Usinga 3×3 Neighborhood[J]. Computer Vision and Image Understanding, 2004, 93(2):195-205.
[42] Toivanen P J. New Geodosic Distance Transforms for Gray-scale Images[J]. PatternRecognition Letters, 1996, 17(5): 437-450.
[43] Gonzalez R C, Woods R E. Digital Images Processing Using MATLAB[M]. NewYork: Addison Wesley Press, 2004: 308-310.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |