基于点特异度和自适应分类策略的眼底图像分割方法

doi:10.11959/j.issn.1000-436x.2015229

摘要/Abstract

摘要：

提出基于点特异度和自适应分类策略的血管分割方法（SSVD,specificity and self-adaptive vessel detection），首先给出点特异度的定义，通过设置高点特异度阈值，实现主血管的提取，然后由多主体进行自适应像素分类，将每个未确定像素作为一个Agent，在多尺度点特异度阈值范围内，根据邻域Agent状态修订自身状态，逐步完成对像素的分类，最后通过多窗口去噪对噪音进行滤除完成对图像血管结构的分割。将SSVD方法应用到DRIVE数据库眼底图像的血管分割中，实验结果表明该方法要比现有其他方法具有更高的准确度和效率。

关键词: 点特异度, 自适应分类策略, 多尺度阈值, 多窗口去噪

Abstract:

A new vessel segmentation method called specificity and self-adaptive vessel detection(SSVD)was proposed based on pixel specificity and self-adaptive classification strategy,in the beginning pixel specificity was defined,by setting a higher pixel specificity threshold,the main vessel skeleton was extracted; then self-adaptive classification process was implemented,and each of the remaining undetermined pixels acted as an Agent,within a multi-scale threshold range,Agent revised its own status according to the status of its neighbor,so as to complete the classification of the pixels; finally the noise was removed by multi-window noise filtering method.By testing SSVD on DRIVE database,the experiment shows that it is more accurate and efficient than state-of-the-art methods.

Key words: pixel specificity, self-adaptive classification strategy;, multi-scale threshold, multi-window noise filtering

姜平,窦全胜. 基于点特异度和自适应分类策略的眼底图像分割方法[J]. 通信学报, 2015, 36(8): 161-170.

Ping JIANG,Quan-sheng DOU. Vessel segmentation of retinal image based on pixel specificity and self-adaptive classification strategy[J]. Journal on Communications, 2015, 36(8): 161-170.

图/表 13

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

表1

图11

表2

参考文献 31

[1]	WASAN B , CERUTTI A , FORD S ,et al. Vascular network changes in the retina with age and hypertension[J]. Journal of Hypertension, 1995,13(12): 1724-1728.
[2]	WONG T Y , MCLNTOSH R . Hypertensive retinopathy signs as risk indicators of cardiovascular morbidity and mortality[J]. British Medical Bulletin, 2005,73(1): 57-70.
[3]	SUSSMAN E J , TSIARAS W G , SOPER K A . Diagnosis of diabetic eye disease[J]. The Journal of the American Medical Association, 1982,247(23): 3231-3234.
[4]	WONG T Y , KLEIN R , COUPER D J . Retinal microvascular abnormalities and incident stroke:the atherosclerosis risk in communities study[J]. The Lancet, 2001,358(9288): 1134-1140.
[5]	WONG T Y , KLEIN R , SHARRETT A R . Retinal arteriolar diameter and risk for hypertension[J]. Annals of Internal Medicine, 2004,140(4): 248-255.
[6]	STAAL J , ABRAMOFF M , NIEMEIJER M ,et al. Ridge-based vessel segmentation in color images of the retina[J]. IEEE Transactions on Medical Imaging, 2004,23: 501-509.
[7]	SOARES J V B , LEANDRO J J G , CESAR R M . Retinal vessel segmentation using the 2D gabor wavelet and supervised classification[J]. IEEE Trans Med Imag, 2006. 1214-1222.
[8]	RICCI E , PERFETTI R . Retinal blood vessel segmentation using line operators and support vector classification[J]. IEEE Transactions on Medical Imaging, 2007,26: 1357-1365.
[9]	LIU J , TANG Y Y . Adaptive image segmentation with distributed behavior-based agents[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1999,21(6): 544-551.
[10]	RICHARD N , DOJAT M , GARBAY C . Automated segmentation of human brain MR images using a multi-agent approach[J]. Artificial Intelligence in Medicine, 2004,30(2): 153-176.
[11]	BOVENKAMP E G P , DIJKSTRA J , BOSCH J G . Multi-agent segmentation of IVUS images[J]. Pattern Recognition, 2004,37(4): 647-663.
[12]	窦全胜, 李国江, 史忠植 ,等. 三维网格空间上的自组装模型[J]. 自动化学报, 2012,38(10): 1595-1601. DOU Q S , LI G J , SHI Z Z ,et al. Collective self-assembly model on three-dimensional grid space[J]. Acta Automatica sinica, 2012,38(10): 1595-1601.
[13]	姜平, 窦全胜, 王晶 . 基于局部自适应点特异度阈值的眼底图像血管分割方法研究[J]. 南京大学学报(自然科学版), 2015,51(1): 165-173. JIANG P , DOU Q S , WANG J . Method for vessel segmentation on retinal image based on local adaptive pixel specificity threshold[J]. Journal of Nanjing University(Natural Science), 2015,51(1): 165-173.
[14]	FRANGI A F , NIESSEN W J , VINCKEN K L ,et al. Multiscale Vessel Enhancement Filtering[M]. Springer,Heidelberg,Germany, 1998.
[15]	BUDAI A , BOCK R , MAIER A ,et al. Robust vessel segmentation in fundus images[J]. International Journal of Biomedical Imaging, Germany, 2013.
[16]	ZANA F , KLEIN J . Robust segmentation of vessels from retinal angiography[A]. Proceedings of the International Conference on Digital Signal Processing[C]. 1997. 1087-1090.
[17]	JIANG , MOJON D . Adaptive local thresholding by verification-based multi-threshold probing with application to vessel detection in retinal images[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003,25: 131-137.
[18]	MENDONCA A , CAMPILHO A . Segmentation of retinal blood vessels by combining the detection of centerlines and morphological reconstruction[J]. IEEE Transactions on Medical Imaging, 2006,25: 1200-1213.
[19]	AL-DIRI B , HUNTER A , STEEL D . An active contour model for segmenting and measuring retinal vessels[J]. IEEE Transactions on Medical Imaging, 2009,28: 1488-1497.
[20]	LAM B , GAO Y , LIEW A . General retinal vessel segmentation using regularization-based multiconcavity modeling[J]. IEEE Transactions on Medical Imaging, 2010,29: 1369-1381.
[21]	MIRI M S , MAHLOOJIFAR A . Retinal image analysis using curvelet transform and multistructure elements morphology by reconstruction[J]. IEEE Transactions on Biomedical Engineering, 2011,58: 1183-1192.
[22]	FRAZ M M , BARMAN S A , REMAGNINO P ,et al. An approach to localize the retinal blood vessels using bit planes and centerline detection,Computer methods and programs in biomedicine[J]. Press,Corrected Proof, 2011,108(2): 600-616.
[23]	YOU X , PENG Q , YUAN Y ,et al. Segmentation of retinal blood vessels using the radial projection and semi-supervised approach,Pattern Recognition[J]. 2011,44: 2314-2324.
[24]	NIEMEIJER M , VANGINNEKEN B , ABRAMOFF M . A linking framework for pixel classification based retinal vessel segmentation[A]. SPIE Medical Imaging[C]. 2009.
[25]	SOARES J , LEANDRO J , JNIOR R . Retinal vessel segmentation using the 2-D gabor wavelet and supervised classification[J]. IEEE Transactions on Medical Imaging, 2006,25: 1214-1222.
[26]	STAAL J , ABRAMOFF M , NIEMEIJER M . Ridge-based vessel segmentation in color images of the retina[J]. IEEE Transactions on Medical Imaging, 2004,23: 501-509.
[27]	LUPASCU C A , TEGOLO D , TRUCCO E . Retinal vessel segmentation using adaboost,information technology in biomedicine[J]. IEEE Transactions on, 2010,14: 1267-1274.
[28]	MARTI R , ZWIGGELAAR R , et al . Automatic registration of mammograms based on linear structures[A]. Information Processing in Medical Imaging[C]. Springer Berlin Heidelberg, 2001,2082: 162-168.
[29]	FRAZ M M , REMAGNINO P ,et al. Blood vessel segmentation methodologies in retinal images-a survey[J]. Computer Methods and Programs in Biomedicine[J]. 2012,108: 407-433.
[30]	ESPONA L , CARREIRA M J ,et al. Progress in Pattern Recognition,Image Analysis and Applications[M]. Berlin,Germany: Springer, 2008,5197.
[31]	MENDON?A A M , CAMPILHO A . Segmentation of retinal blood vessels by combining the detection of centerlines and morphological reconstruction[J]. IEEE Transactions on Medical Imaging, 2006,25(9): 1200-1213.

序号	类型	方法	年	Acc	AUC
1		2nd observer	—	0.947 0	N/A
2		文献[16]方法	2001	0.937 7	0.898 4
3		文献[17]方法	2003	0.921 2	0.911 4
4		文献[18]方法	2006	0.945 2	N/A
56	无监督	文献[19]方法	2009	N/A	N/A
56	无监督	文献[20]方法	2010	0.947 2	0.961 4
7		文献[21]方法	2011	0.945 8	N/A
8		文献[22]方法	2011	0.943 0	N/A
9		文献[23]方法	2011	0.943 4	N/A
10		文献[24]方法	2004	0.941 6	0.929 4
11		文献[25]方法	2006	0.946 6	0.961 4
12		文献[26]方法	2004	0.944 1	0.952 0
13	监督	文献[8]方法	2007	0.959 5	0.955 8
14		文献[27]方法	2010	0.959 7	0.956 1
15		文献[28]方法	2011	0.945 2	0.958 8
16		文献[29]方法	2012	0.953 4	0.976 8
17		SSVD方法	2015	0.954 0	0.977 9

算法	运行时间/s
SSVD算法	23.03
文献[30]算法	31.7
文献[31]算法	150
文献[7]算法	180
文献[24]算法	900