复杂背景下的宫颈细胞核分割方法
|
摘要
自动阅片技术采用图像处理方法,在细胞核识别的基础上实现对细胞核DNA含量的准确测量,能够为医生诊断提供辅助。图像分割作为自动阅片系统关键直接影响系统性能。然而,显微镜下细胞核图片存在光照不均、背景阴影等情况。并且不可避免地存在一些血细胞、淋巴细胞、垃圾杂质、成团细胞,这严重降低了分割算法的性能。为解决上述问题,提出了复杂背景下的宫颈细胞核分割方法。该方法首先利用参数自适应的局部阈值法来分割图片,并利用自适应的分割参数调节来处理图片中的复杂情况。即利用局部阈值窗口大小和二值化后轮廓数量的函数关系来自动调节窗口大小。然后根据得到的二值图确定分水岭算法的标记图像,最后利用分水岭算法完成整张图片的分割。实验表明,此方法能适应复杂图像环境,并可以从成团细胞中分割出重叠程度较低的细胞,提高分类器识别细胞核的概率。
Automatic screening technology developed in recent years. It applies image processing,and first recognizes nucleus and then measures the DNA contents accurately,so it can provide auxiliary for a doctor' s diagnosis. Image segmentation is the key technique of automatic screening system which directly determines the performance of the systems. However,the imaging environments under the microscope are complex. One the one hand,uneven illumination,background shading and uneven dyed nucleus exist. On the other hand,there are inevitably blood cells,lymphocytes,garbage,impurities and conglobation cells in cell images. These conditions degrade the performance of image segmentation. In order to solve these problems,we put forward a method to segment cervical nuclei in complex background. This method first employs the local threshold method to segment images. In this procedure we propose a parameter adapting method which adjusts its parameters automatically according to the function of local threshold window size and the binarized outline number. The local threshold method transforms an image into a binary image which is then passed to image corrosion operator to generate a marking image. With the binary image,the watershed algorithm was finally performed to segment the image. The experiment shows that the method can adapt to the complex image environment and separate the cells with lower overlapping nuclei images.
Automatic screening technology developed in recent years. It applies image processing,and first recognizes nucleus and then measures the DNA contents accurately,so it can provide auxiliary for a doctor' s diagnosis. Image segmentation is the key technique of automatic screening system which directly determines the performance of the systems. However,the imaging environments under the microscope are complex. One the one hand,uneven illumination,background shading and uneven dyed nucleus exist. On the other hand,there are inevitably blood cells,lymphocytes,garbage,impurities and conglobation cells in cell images. These conditions degrade the performance of image segmentation. In order to solve these problems,we put forward a method to segment cervical nuclei in complex background. This method first employs the local threshold method to segment images. In this procedure we propose a parameter adapting method which adjusts its parameters automatically according to the function of local threshold window size and the binarized outline number. The local threshold method transforms an image into a binary image which is then passed to image corrosion operator to generate a marking image. With the binary image,the watershed algorithm was finally performed to segment the image. The experiment shows that the method can adapt to the complex image environment and separate the cells with lower overlapping nuclei images.
引文
[1]吴一全,孟天亮,吴诗婳.图像阈值分割方法研究进展20年(1994-2014)[J].数据采集与处理,2015,30(1):1.
[2]XING F,LIN Y.Robust Nucleus/Cell Detection and Segmentation in Digital Pathology and Microscopy Images:A Comprehensive Review[J].IEEE Reviews in Biomedical Engineering,2016,9.
[3]申铉京,刘翔,陈海鹏.基于多阈值Otsu准则的阈值分割快速计算[J].电子与信息学报,2017,v.39(01):144.
[4]ZENG Z,CHEN S,TANG S,et al.Unsupervised Segmentation of Cell Nuclei in Cervical Smear Images Using Active Contour with A-daptive Local Region Fitting Energy Modelling[C]//International Conference on Biomedical Engineering and Informatics.2015:250.
[5]刘占.基于局部期望阈值分割的图像边缘检测算法[J].计算机与现代化,2016(8):52.
[6]RANEFALL P,SADANANDAN S K,WHLBY C.Fast Adaptive Local Thresholding Based on Ellipse Fit[C]//The IEEE International Symposium on Biomedical Imaging.IEEE,2016.
[7]JI X,LI Y,CHENG J,et al.Cell Image Segmentation Based on an Improved Watershed Algorithm[C]//Image and Signal Processing(CISP),2015 8th International Congress on.IEEE,2015:433.
[8]GEETHA P K,NIDHYA R,DINESH Kumar A,et al.Cell Segmentation and NC Ratio Analysis for Biopsy Images Using Marker Controlled Watershed Algorithm[C]//International Conference on Green Computing Communication and Electrical Engineering.IEEE,2014:1.
[9]HUSAIN R A,ZAYED A S,AHMED W M,et al.Image Segmentation with Improved Watershed Algorithm Using Radial Bases Function Neural Networks[C]//Sciences and Techniques of Automatic Control and Computer Engineering(STA),2015 16th International Conference on.IEEE,2015:121.
[10]ZHANPENG H,QI Z,SHIZHONG J,et al.Medical Image Segmentation Based on the Watersheds and Regions Merging[C]//Information Science and Control Engineering(ICISCE),2016 3rd International Conference on.IEEE,2016:1011.
[11]杨辉华,赵玲玲,潘细朋,等.基于水平集和凹点区域检测的粘连细胞分割方法[J].北京邮电大学学报,2016,39(6):11.
[2]XING F,LIN Y.Robust Nucleus/Cell Detection and Segmentation in Digital Pathology and Microscopy Images:A Comprehensive Review[J].IEEE Reviews in Biomedical Engineering,2016,9.
[3]申铉京,刘翔,陈海鹏.基于多阈值Otsu准则的阈值分割快速计算[J].电子与信息学报,2017,v.39(01):144.
[4]ZENG Z,CHEN S,TANG S,et al.Unsupervised Segmentation of Cell Nuclei in Cervical Smear Images Using Active Contour with A-daptive Local Region Fitting Energy Modelling[C]//International Conference on Biomedical Engineering and Informatics.2015:250.
[5]刘占.基于局部期望阈值分割的图像边缘检测算法[J].计算机与现代化,2016(8):52.
[6]RANEFALL P,SADANANDAN S K,WHLBY C.Fast Adaptive Local Thresholding Based on Ellipse Fit[C]//The IEEE International Symposium on Biomedical Imaging.IEEE,2016.
[7]JI X,LI Y,CHENG J,et al.Cell Image Segmentation Based on an Improved Watershed Algorithm[C]//Image and Signal Processing(CISP),2015 8th International Congress on.IEEE,2015:433.
[8]GEETHA P K,NIDHYA R,DINESH Kumar A,et al.Cell Segmentation and NC Ratio Analysis for Biopsy Images Using Marker Controlled Watershed Algorithm[C]//International Conference on Green Computing Communication and Electrical Engineering.IEEE,2014:1.
[9]HUSAIN R A,ZAYED A S,AHMED W M,et al.Image Segmentation with Improved Watershed Algorithm Using Radial Bases Function Neural Networks[C]//Sciences and Techniques of Automatic Control and Computer Engineering(STA),2015 16th International Conference on.IEEE,2015:121.
[10]ZHANPENG H,QI Z,SHIZHONG J,et al.Medical Image Segmentation Based on the Watersheds and Regions Merging[C]//Information Science and Control Engineering(ICISCE),2016 3rd International Conference on.IEEE,2016:1011.
[11]杨辉华,赵玲玲,潘细朋,等.基于水平集和凹点区域检测的粘连细胞分割方法[J].北京邮电大学学报,2016,39(6):11.