结合局部熵和鲁棒主成分分析的眼底图像硬性渗出物检测方法

doi:10.11772/j.issn.1001-9081.2019010208

计算机应用 ›› 2019, Vol. 39 ›› Issue (7): 2134-2140.DOI: 10.11772/j.issn.1001-9081.2019010208

• 虚拟现实与多媒体计算 • 上一篇下一篇

结合局部熵和鲁棒主成分分析的眼底图像硬性渗出物检测方法

陈莉^1,2, 陈晓云²

1. 福建医科大学基础医学院, 福州 350108;
2. 福州大学数学与计算机科学学院, 福州 350116

收稿日期:2019-01-28 修回日期:2019-03-09 出版日期:2019-07-10 发布日期:2019-07-15
通讯作者: 陈晓云
作者简介:陈莉(1980-),女,福建福清人,讲师,硕士,主要研究方向:模式识别、医学图像处理;陈晓云(1970-),女,福建福州人,教授,博士,主要研究方向:机器学习、模式识别。
基金资助:
国家自然科学基金资助项目（71273053，11571074）；福建省自然科学基金资助项目（2018J01666）。

Detection method of hard exudates in fundus images by combining local entropy and robust principal components analysis

CHEN Li^1,2, CHEN Xiaoyun²

1. School of Basic Medical Sciences, Fujian Medical University, Fuzhou Fujian 350108, China;
2. College of Mathematics and Computer Science, Fuzhou University, Fuzhou Fujian 350116, China

Received:2019-01-28 Revised:2019-03-09 Online:2019-07-10 Published:2019-07-15
Supported by:
This work is partially supported by the National Natural Science Foundation of China (71273053, 11571074), the Natural Science Foundation of Fujian Province (2018J01666).

摘要/Abstract

摘要：

针对眼科医生诊断眼底图像工作耗时且易出错的问题，提出一种无监督的眼底图像硬性渗出物检测方法。首先，通过形态学的背景估计方法去除血管、暗病变区域和视盘；然后，以图像亮度通道为初始图像，利用硬性渗出物在眼底图像中的局部性和稀疏性，结合局部熵和鲁棒主成分分析方法分解得到低秩矩阵和稀疏矩阵；最后，归一化稀疏矩阵得到硬性渗出物区域。实验结果显示，在e-ophtha EX和DIARETDB1公开数据库上，所提方法在病灶水平上灵敏性为91.13%和特异性为90%，在图像水平上准确率为99.03%，平均运行时间0.5 s；与支持向量机（SVM）和K-means方法相比灵敏性高且耗时少。

关键词: 硬性渗出物, 鲁棒主成分分析, 局部熵, 背景估计, 彩色眼底图像

Abstract:

To solve the time-consuming and error-prone problem in the diagnosis of fundus images by the ophthalmologists, an unsupervised automatic detection method for hard exudates in fundus images was proposed. Firstly, the blood vessels, dark lesion regions and optic disc were removed by using morphological background estimation in preprocessing phase. Then, with the image luminosity channel taken as the initial image, the low rank matrix and sparse matrix were obtained by combining local entropy and Robust Principal Components Analysis (RPCA) based on the locality and sparsity of hard exudates in fundus images. Finally, the hard exudates regions were obtained by the normalized sparse matrix. The performance of the proposed method was tested on the fundus images databases e-ophtha EX and DIARETDB1. The experimental results show that the proposed method can achieve 91.13% of sensitivity and 90% of specificity in the lesional level and 99.03% of accuracy in the image level and 0.5 s of average running time. It can be seen that the proposed method has higher sensitivity and shorter running time compared with Support Vector Machine (SVM) method and K-means method.

Key words: hard exudate, Robust Principal Components Analysis (RPCA), local entropy, background estimation, color fundus image

中图分类号:

TP391.413

陈莉, 陈晓云. 结合局部熵和鲁棒主成分分析的眼底图像硬性渗出物检测方法[J]. 计算机应用, 2019, 39(7): 2134-2140.

CHEN Li, CHEN Xiaoyun. Detection method of hard exudates in fundus images by combining local entropy and robust principal components analysis[J]. Journal of Computer Applications, 2019, 39(7): 2134-2140.

参考文献

[1] 肖志涛,王雯,耿磊,等.基于背景估计和SVM分类器的眼底图像硬性渗出物检测方法[J].中国生物医学工程学报,2015,34(6):720-728.(XIAO Z T, WANG W, GENG L, et al. Hard exudates detection method based on background-estimation and SVM classifier[J]. Chinese Journal of Biomedical Engineering, 2015, 34(6):720-728.)
[2] WELFER D, SCHARCANSKI J, MARINHO D R. A coarse-to-fine strategy for automatically detecting exudates in color eye fundus images[J]. Computerized Medical Imaging Graphics, 2010, 34(3):228-235.
[3] WALTER T, KLEIN J C, MASSIN P, et al. A contribution of image processing to the diagnosis of diabetic retinopathy detection of exudates in color fundus images of the human retina[J]. IEEE Transactions on Medical Imaging, 2002, 21(10):1236-1243.
[4] QI F C, LI G, ZHENG S B. Automatic exudate detection in color fundus images[J]. Digital TV and Wireless Multimedia Communication, 2017, 685:155-165.
[5] SANTHI D, MANIMEGALAI D, PARVATHI S, et al. Segmentation and classification of bright lesions to diagnose diabetic retinopathy in retinal images[J]. Biomedizinische Technik (BIOMED TECH), 2016, 61(4):443-453.
[6] 曹新容,林嘉雯,薛岚燕,等.邻域约束模型的眼底图像硬性渗出聚类检测方法[J].计算机辅助设计与图形学学报,2018,30(11):2093-2100.(CAO X R, LIN J W, XUE L Y, et al. Clustering detection method of hard exudates in fundus image based on neighborhood constraint model[J]. Journal of Computer-Aided Design and Computer Graphics, 2018, 30(11):2093-2100.)
[7] SOPHARAK A, UYYANONVARA B, BARMAN S, et al. Comparative analysis of automatic exudate detection algorithms[C]//WCE 2010:Proceedings of the 2010 World Congress on Engineering. London:[s.n.], 2010:738-741.
[8] BIYANI R S, PATRE B M. A clustering approach for exudates detection in screening of diabetic retinopathy[C]//Proceedings of the 2016 International Conference on Signal and Information Processing. Washington, DC:IEEE Computer Society, 2016:1-5.
[9] ASHA P R, KARPAGAVALLI S. Diabetic retinal exudates detection using machine learning techniques[C]//Proceedings of the 2015 International Conference on Advanced Computing and Communication System. Piscataway, NJ:IEEE, 2015:1-5.
[10] ADEM K. Exudate detection for diabetic retinopathy with circular Hough transformation and convolutional neural networks[J]. Expert Systems with Applications, 2018, 11:289-295.
[11] PARHAM K, PASSOS J L A, TIAGO C, et al. Exudate detection in fundus images using deeply-learnable features[J]. Computers in Biology and Medicine, 2018, 104:62-69.
[12] AMIN J, SHARIF M, YASMIN M, et al. A method for the detection and classification of diabetic retinopathy using structural predictors of bright lesions[J]. Journal of Computational Science, 2017, 19:153-164.
[13] WRIGHT J, GANESH A, RAO S, et al. Robust principal component analysis:exact recovery of corrupted low-rank matrices via convex optimization[EB/OL].[2018-12-10]. https://arxiv.org/pdf/0905.0233v1.pdf.
[14] XIAO L, FANG B, LIU L H, et al. Extracting sparse error of robust PCA for face recognition in the presence of varying illumination and occlusion[J]. Pattern Recognition, 2014, 47(2):495-508.
[15] YAN J C, ZHU M Y, LIU H X, et al. Visual saliency detection via sparsity pursuit[J]. IEEE Signal Processing Letters, 2010, 17(8):739-742.
[16] 仓园园,孙玉宝,刘青山.基于分层鲁棒主成分分析的运动目标检测[J].计算机辅助设计与图形学学报,2014,23(4):537-544.(CANG Y Y, SUN Y B, LIU Q S. Moving object detection based on hierarchical robust principal component analysis[J]. Journal of Computer-Aided Design and Computer Graphics, 2014, 23(4):537-544.)
[17] CANDÉS E J, LI X, MA Y, et al. Robust principal component analysis?[J]. Journal of the ACM, 2011, 58(3):Article No. 11.
[18] AZIZI S, SAMAVI S, MOHREKESH M, et al. Cascaded transform space watermarking based on analysis of local entropy variation[C]//Proceedings of the 2013 International Conference on Multimedia and Expo Workshops. Piscataway, NJ:IEEE, 2013:1-6.
[19] 付晓薇,代芸,陈黎,等.基于局部熵的量子衍生医学超声图像去斑[J].电子与信息学报,2015,37(3):560-566.(FU X W, DAI Y, CHEN L, et al. Quantum-inspired despeckling of medical ultrasound images based on local entropy[J]. Journal of Electronics and Information Technology, 2015, 37(3):560-566.)
[20] MASSIN P, CHABOUIS A, ERGINAY A, et al. OPHDIAT^©:A telemedical network screening system for diabetic retinopathy in the Île-de-France[J]. Diabetes and Metabolism, 2008, 34(3):227-234.
[21] DECENCIERE E, CAZUGUEL G, ZHANG X W, et al. TeleOphta:machine learning and image processing methods for teleophthalmology[J]. Innovation and Research in BioMedical Engineering, 2013, 34(2):196-203.
[22] ZHANG X W, THIBAULT G, DECENCIÈRE E, et al. Exudate detection in color retinal images for mass screening of diabetic retinopathy[J]. Medical Image Analysis, 2014, 18(7):1026-1043.

结合局部熵和鲁棒主成分分析的眼底图像硬性渗出物检测方法

Detection method of hard exudates in fundus images by combining local entropy and robust principal components analysis

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 7

编辑推荐

Metrics

[1]	王心, 朱浩华, 刘光灿. 卷积鲁棒主成分分析[J]. 计算机应用, 2021, 41(5): 1314-1318.
[2]	陈利霞, 班颖, 王学文. 基于张量核范数与3D全变分的背景减除[J]. 计算机应用, 2020, 40(9): 2737-2742.
[3]	王海鹏, 降爱莲, 李鹏翔. 牛顿-软阈值迭代鲁棒主成分分析算法[J]. 计算机应用, 2020, 40(11): 3133-3138.
[4]	孙莞格, 夏克文, 兰璞. 正则化的加权不完全鲁棒主成分分析方法及其在无线传感器网络节点轨迹拟合中的应用[J]. 计算机应用, 2018, 38(6): 1709-1714.
[5]	田继伟, 王布宏, 尚福特. 基于鲁棒主成分分析的智能电网虚假数据注入攻击[J]. 计算机应用, 2017, 37(7): 1943-1947.
[6]	史加荣, 郑秀云, 杨威. 不完全鲁棒主成分分析的正则化方法及其在背景建模中的应用[J]. 计算机应用, 2015, 35(10): 2824-2827.
[7]	范海菊冯云芝王涛冯乃勤. 离焦模糊图像模糊半径检测的新方法[J]. 计算机应用, 2012, 32(07): 1875-1878.