尿沉渣图像有形成分分割与两类细胞识别
|
摘要
自动尿沉渣分类仪器对于临床尿检具有重要的意义。它解决了长期以来传统的通过人工对尿沉渣涂片进行目镜检查的操作过程中存在的受技术水平,视觉上的偏差影响,工作效率低,无法给出定量分析结果的弊端,把医务人员从繁重的重复性劳动中解放出来。排除各种由于主观因素而产生的误差,实现对图像的快速精确定量分析,提高诊断的准确性。同时,数字化的检验结果和图样可以方便的实现图像,数据的远程调用,为远程医疗与疾病的会诊等工作提供便利。
本文在数字图像处理与模式识别的理论基础上,通过深入的研究和大量的实验,提出了一个针对复杂尿沉渣显微图像中各种有形成分分割定位于两类有成成分即精子,管型细胞识别的系统方法。本系统能够将数字显微图像中可以为诊断提供依据的各类有形成分:白细胞,红细胞,上皮细胞,管型,精子,结晶,霉菌等从背景中很好的分割出来,并通过决策树分类器识别出其中的管型细胞和精子。
首先,在图像分割方面,本文充分利用8方向灰度信息平衡光照不均等因素,得到标准差梯度图像,采用基于标准差梯度图的直方图统计信息的双阈值分割算法对图像进行二值化,并对初步分割后的图像根据局部灰度图像灰度均一性强度有选择性的选取canny操作的阈值,随后基于canny图,通过形态学闭操作和5X5邻域的图像边缘链接操作,对初步二值图像进行紧致。最后针对对多细胞粘连,细胞团的情况,基于距离图与部分先验知识进行多细胞分割,定位细胞团中的每个细胞。
在特征选取方面,利用管型细胞均匀圆柱体,精子头尾两部分及尾部呈粗细均匀的管状结构等典型特征分别设计提取了多个形态特征参数,并对管型设计了简单的纹理特征。
在分类器设计上面,为了提高识别器的效率,降低其复杂度,选取了基于二叉决策树的分类器,结合形态特征和纹理特征做为分类属性,识别两类有形成分。
经临床大量尿液样本验证表明,本文设计的系统使得尿沉渣图像中各有形成分分割完整,定位准确,为下一步识别垫定了良好的基础。在粗筛定位上,管型和精子的识别率也达到了较好的水平。
Automatic urine sediment classifier has significant influence on clinic urine analy-sis. As compared to the traditional manual way which is limited by the technical level, deviation error based on vision and low efficience, it can relieve the doctors of their hard, time consuming manual work and avoid diagnostic error caused by subjectivism. Moreover, it provides quantitative analysis and high efficience. Meanwhile, digital re-sults and pictures are convient for long-distance transfer which is important for long-distance medical treatment and consultation.
Base on theories of digital image processing and pattern recogniton and some ex-periment, we develop a strategy for objects detection and recognition of two of them, which are cast and sperm. Our system is able to segment various sediments which are useful for diagnostic analysis from the backgroud of digital microscopic image. These sediments include white blood cell, red blood cell, epithclium, cast, sperm, crystal and mycete. The classifier of system can recognize cast and sperm from other objects.
In image segmentation stage, first of all, based on the statistic information of stan-dard difference gradient image’s histogram, we employ an bi-thresholding method. Second, a refining strategy adopting the combined canny and gray image information is selectivly applied to some lacal part of the previous binary image.At last, in terms of cell mass, in the light of priori knowledge of blood cells, we exlpore a method making use of distantce transform image information to locate each of the blood cells in the mass and thereby segment them from the mass.
In feature selcetion, we extract some shape features according cast and sperm’s physic characteristics. As for cast, we additionally design a texture feature.
In classifier designing, classifiers based on dicision tree are developed for high ef-ficiency and simpleness. The classifier for cast combines shape features with texture features.
A large number experimental results show that our strategy can extract objects from background neatly and precisely. Meanwhile, it can also achieve satisfactort recognition rates.
本文在数字图像处理与模式识别的理论基础上,通过深入的研究和大量的实验,提出了一个针对复杂尿沉渣显微图像中各种有形成分分割定位于两类有成成分即精子,管型细胞识别的系统方法。本系统能够将数字显微图像中可以为诊断提供依据的各类有形成分:白细胞,红细胞,上皮细胞,管型,精子,结晶,霉菌等从背景中很好的分割出来,并通过决策树分类器识别出其中的管型细胞和精子。
首先,在图像分割方面,本文充分利用8方向灰度信息平衡光照不均等因素,得到标准差梯度图像,采用基于标准差梯度图的直方图统计信息的双阈值分割算法对图像进行二值化,并对初步分割后的图像根据局部灰度图像灰度均一性强度有选择性的选取canny操作的阈值,随后基于canny图,通过形态学闭操作和5X5邻域的图像边缘链接操作,对初步二值图像进行紧致。最后针对对多细胞粘连,细胞团的情况,基于距离图与部分先验知识进行多细胞分割,定位细胞团中的每个细胞。
在特征选取方面,利用管型细胞均匀圆柱体,精子头尾两部分及尾部呈粗细均匀的管状结构等典型特征分别设计提取了多个形态特征参数,并对管型设计了简单的纹理特征。
在分类器设计上面,为了提高识别器的效率,降低其复杂度,选取了基于二叉决策树的分类器,结合形态特征和纹理特征做为分类属性,识别两类有形成分。
经临床大量尿液样本验证表明,本文设计的系统使得尿沉渣图像中各有形成分分割完整,定位准确,为下一步识别垫定了良好的基础。在粗筛定位上,管型和精子的识别率也达到了较好的水平。
Automatic urine sediment classifier has significant influence on clinic urine analy-sis. As compared to the traditional manual way which is limited by the technical level, deviation error based on vision and low efficience, it can relieve the doctors of their hard, time consuming manual work and avoid diagnostic error caused by subjectivism. Moreover, it provides quantitative analysis and high efficience. Meanwhile, digital re-sults and pictures are convient for long-distance transfer which is important for long-distance medical treatment and consultation.
Base on theories of digital image processing and pattern recogniton and some ex-periment, we develop a strategy for objects detection and recognition of two of them, which are cast and sperm. Our system is able to segment various sediments which are useful for diagnostic analysis from the backgroud of digital microscopic image. These sediments include white blood cell, red blood cell, epithclium, cast, sperm, crystal and mycete. The classifier of system can recognize cast and sperm from other objects.
In image segmentation stage, first of all, based on the statistic information of stan-dard difference gradient image’s histogram, we employ an bi-thresholding method. Second, a refining strategy adopting the combined canny and gray image information is selectivly applied to some lacal part of the previous binary image.At last, in terms of cell mass, in the light of priori knowledge of blood cells, we exlpore a method making use of distantce transform image information to locate each of the blood cells in the mass and thereby segment them from the mass.
In feature selcetion, we extract some shape features according cast and sperm’s physic characteristics. As for cast, we additionally design a texture feature.
In classifier designing, classifiers based on dicision tree are developed for high ef-ficiency and simpleness. The classifier for cast combines shape features with texture features.
A large number experimental results show that our strategy can extract objects from background neatly and precisely. Meanwhile, it can also achieve satisfactort recognition rates.
引文
[1]顾可梁.尿有形成分的识别与检查方法的选择[J].中华检验医学杂志, 2005, 28(6) : 25-29.
[2]叶应歧.计算机技术在医学诊断和研究中的应用[M].北京:中国中医药出版社,2008: 37-56.
[3] Tzu-I Chien. Urine sediment examination: A comparison of automated urinalysis systems and manual microscopy[J]. Clinica Chimica Acta 2007, vol(384):28–34.
[4] M. Ranzato.Automatic recognition of biological particles in microscopic images[J]. Pattern Recognition Letters. 2007, vol(28):31–39.
[5]沈美丽.尿沉渣有形成分自动分类系统研究[D].长春理工大学博士学位论文,2008.
[6] Pornvaree Lamchiagdhase. Urine sediment examination: A comparison between the manual method and the iQ200 automated urine microscopy analyzer[J]. Clinica Chimica Acta 2005,vol(358): 167–174.
[7]高月红.图像分割算法的研究[J].科技信息, 2009,(27).
[8]王云霞.医学细胞显微图像分割与识别技术的研究[D].黑龙江大学硕士学位论文, 2008.
[9] Y.J.Zhang.Evaluation and comparison of different segmentation algorithms.Pattern recogni-tion letters[J],1997,18(6):963-974.
[10]韩思奇,王蕾.图像分割的阈值法综述[J].系统工程与电子技术,2002,24(6): 91-94.
[11] Rosin.P.L.Unimodal thresholding[J].Pattern recognition,2001,34(11):2083-2096.
[12] Yan Feixiang , Zhang Hong , Kube CR. A multistage adaptive thresholding method [J ] . Pat-tern recognition letters , 2005 , 26 (8) :1183 - 1191.
[13] Savakis AE. Adaptive document image thresholding using foreground and background clus-tering [C] . Chicago : Springer2Verlag , 1998.37 - 42.
[14]章毓晋.图像分割[M].北京:科学出版社,2001.
[15] VINCENTL,SOILLE P. Watersheds in digital space:An efficient algorithms based on immer-sions in ulation[J]. IEEE Transactions on Pattern Analysis and Machine Intellgence, 1991,13(6):583-598.
[16]张新峰,沈兰荪.图像分割技术研究[J].电路与系统学报,2004,9(2):92-99.
[17]王选贺.基于尿沉渣显微图像的模式识别[D].吉林大学硕士学位论文, 2007.
[18]何斌等. Visual C++数字图像处理[M].北京:人民邮电出版社,2001.
[19]唐艳平,张灿龙.基于形态学和Snake模型的尿沉渣提取[J].桂林电子科技大学学报,2007,12Vol.27(6).
[20] O she r SJ , FedkiwRP. Level Set Methods and Dynamic Implic it Surfaces [M ]. New York:Springer2Verlag Press, 2003.
[21] S.Nomura,K.Yamanaka,O.Katai.New adaptive methods applied to printed word image bina-rization.Proceedings of the fourth IASTED International conference signal and image processing[M].USA.2002.USA:IASTED,2002,8:12-14,288-293.
[22] N. OTSU, A threshold selection method from gray-level histogram[J]. IEEE Trans. On Sys-tem, Man and Cybernetics, Vol.9, No1:62-66,1979.
[23]苏传朋.尿沉渣显微图像有形成分自动分割算法研究[D].浙江大学硕士学位论文2006.
[24]何家英等.基于数学形态学和Canny算子的边缘提取方法[J].计算机应用2007,6:477-478.
[25]徐达丽等.基于链码技术的距离变换改进算法[J].计算机工程与应用. 2009 45 (25): 176-178.
[26] Quan Pang,Cuirong Yang,Yingle Fan,Yu Chen. Overlapped Cell Image Segmentation Based on Distance Transform[J]. Proceedings of the 6th World Congress on Intelligent Control and Automation, June 21 - 23, 2006, Dalian, China.
[27]李勇明.尿沉渣图像自动识别算法的研究[D].重庆大学博士学位论文,2007.
[28]孙即祥等.现代模式识别[M].长沙:国防科技大学出版社,2001.5:23-345.
[29]边肇祺,张学工.模式识别第二版[M].北京:清华大学出版社.1999,12:33-96.
[30] Simon Haykin(叶世伟等译).神经网络原理[M].北京:机械工业出版社,2004
[31] Olivia R. Liu Sheng, Chih-Ping Wei, Paul Jen-Hwa Hu, Namsik Chang.Automated learning of patient image retrieval knowledge: neural networks versus inductive decision trees[J]. Deci-sion Support Systems, Volume 30, Issue 2, 27 December 2000, Pages 105-124.
[32]胡记兵.基于决策树的组合分类器的构建和部署[D].浙江大学硕士学位论文2008.
[33] Y.P. Ginoris, A.L. Amaral, A. Nicolau, M.A.Z. Coelho, E.C. Ferreira.Recognition of protozoa and metazoa using image analysis tools, discriminant analysis, neural networks and decision trees[J]. Analytica Chimica Acta. 2007(595): 160-169.
[34]顾可梁.尿沉渣管型检查.临床检验杂志[J]. 2001,(1):23-26.
[35]李雯王文华,刘灿.三种尿沉渣检测法在尿有形成分检测中的临床应用[J].国际检验医学杂志.2007,(28):342-344.
[36]彩色细胞显微图像自动识别技术的研究[D].山东大学硕士学位论文.2008
[2]叶应歧.计算机技术在医学诊断和研究中的应用[M].北京:中国中医药出版社,2008: 37-56.
[3] Tzu-I Chien. Urine sediment examination: A comparison of automated urinalysis systems and manual microscopy[J]. Clinica Chimica Acta 2007, vol(384):28–34.
[4] M. Ranzato.Automatic recognition of biological particles in microscopic images[J]. Pattern Recognition Letters. 2007, vol(28):31–39.
[5]沈美丽.尿沉渣有形成分自动分类系统研究[D].长春理工大学博士学位论文,2008.
[6] Pornvaree Lamchiagdhase. Urine sediment examination: A comparison between the manual method and the iQ200 automated urine microscopy analyzer[J]. Clinica Chimica Acta 2005,vol(358): 167–174.
[7]高月红.图像分割算法的研究[J].科技信息, 2009,(27).
[8]王云霞.医学细胞显微图像分割与识别技术的研究[D].黑龙江大学硕士学位论文, 2008.
[9] Y.J.Zhang.Evaluation and comparison of different segmentation algorithms.Pattern recogni-tion letters[J],1997,18(6):963-974.
[10]韩思奇,王蕾.图像分割的阈值法综述[J].系统工程与电子技术,2002,24(6): 91-94.
[11] Rosin.P.L.Unimodal thresholding[J].Pattern recognition,2001,34(11):2083-2096.
[12] Yan Feixiang , Zhang Hong , Kube CR. A multistage adaptive thresholding method [J ] . Pat-tern recognition letters , 2005 , 26 (8) :1183 - 1191.
[13] Savakis AE. Adaptive document image thresholding using foreground and background clus-tering [C] . Chicago : Springer2Verlag , 1998.37 - 42.
[14]章毓晋.图像分割[M].北京:科学出版社,2001.
[15] VINCENTL,SOILLE P. Watersheds in digital space:An efficient algorithms based on immer-sions in ulation[J]. IEEE Transactions on Pattern Analysis and Machine Intellgence, 1991,13(6):583-598.
[16]张新峰,沈兰荪.图像分割技术研究[J].电路与系统学报,2004,9(2):92-99.
[17]王选贺.基于尿沉渣显微图像的模式识别[D].吉林大学硕士学位论文, 2007.
[18]何斌等. Visual C++数字图像处理[M].北京:人民邮电出版社,2001.
[19]唐艳平,张灿龙.基于形态学和Snake模型的尿沉渣提取[J].桂林电子科技大学学报,2007,12Vol.27(6).
[20] O she r SJ , FedkiwRP. Level Set Methods and Dynamic Implic it Surfaces [M ]. New York:Springer2Verlag Press, 2003.
[21] S.Nomura,K.Yamanaka,O.Katai.New adaptive methods applied to printed word image bina-rization.Proceedings of the fourth IASTED International conference signal and image processing[M].USA.2002.USA:IASTED,2002,8:12-14,288-293.
[22] N. OTSU, A threshold selection method from gray-level histogram[J]. IEEE Trans. On Sys-tem, Man and Cybernetics, Vol.9, No1:62-66,1979.
[23]苏传朋.尿沉渣显微图像有形成分自动分割算法研究[D].浙江大学硕士学位论文2006.
[24]何家英等.基于数学形态学和Canny算子的边缘提取方法[J].计算机应用2007,6:477-478.
[25]徐达丽等.基于链码技术的距离变换改进算法[J].计算机工程与应用. 2009 45 (25): 176-178.
[26] Quan Pang,Cuirong Yang,Yingle Fan,Yu Chen. Overlapped Cell Image Segmentation Based on Distance Transform[J]. Proceedings of the 6th World Congress on Intelligent Control and Automation, June 21 - 23, 2006, Dalian, China.
[27]李勇明.尿沉渣图像自动识别算法的研究[D].重庆大学博士学位论文,2007.
[28]孙即祥等.现代模式识别[M].长沙:国防科技大学出版社,2001.5:23-345.
[29]边肇祺,张学工.模式识别第二版[M].北京:清华大学出版社.1999,12:33-96.
[30] Simon Haykin(叶世伟等译).神经网络原理[M].北京:机械工业出版社,2004
[31] Olivia R. Liu Sheng, Chih-Ping Wei, Paul Jen-Hwa Hu, Namsik Chang.Automated learning of patient image retrieval knowledge: neural networks versus inductive decision trees[J]. Deci-sion Support Systems, Volume 30, Issue 2, 27 December 2000, Pages 105-124.
[32]胡记兵.基于决策树的组合分类器的构建和部署[D].浙江大学硕士学位论文2008.
[33] Y.P. Ginoris, A.L. Amaral, A. Nicolau, M.A.Z. Coelho, E.C. Ferreira.Recognition of protozoa and metazoa using image analysis tools, discriminant analysis, neural networks and decision trees[J]. Analytica Chimica Acta. 2007(595): 160-169.
[34]顾可梁.尿沉渣管型检查.临床检验杂志[J]. 2001,(1):23-26.
[35]李雯王文华,刘灿.三种尿沉渣检测法在尿有形成分检测中的临床应用[J].国际检验医学杂志.2007,(28):342-344.
[36]彩色细胞显微图像自动识别技术的研究[D].山东大学硕士学位论文.2008