一种基于改进U形网络的眼底图像视网膜新生血管检测方法  

A Method of Retinal Neovascularization Detection on Retinal Image Based on Improved U-net

作　　者：邹北骥[2,3] 易博松 刘晴 ZOU Beiji;YI Bosong;LIU Qing(School of Automation,Central South University,Changsha 410083,China;School of Computing,Central South University,Changsha 410083,China;Hunan Machine Vision and Intelligent Medical Research Center,Changsha 410083,China)

机构地区：[1]中南大学自动化学院,湖南长沙410083 [2]中南大学计算机学院,湖南长沙410083 [3]湖南省机器视觉与智慧医疗工程技术研究中心,湖南长沙410083

出　　处：《湖南大学学报（自然科学版）》2021年第4期19-25,共7页Journal of Hunan University:Natural Sciences

基　　金：国家自然科学基金资助项目(61573380,61702559);国家重大科技专项(2018AAA0102100)。

摘　　要：糖尿病性视网膜病变(简称糖网病)是主要的致盲眼疾病之一,视网膜新生血管的出现是糖网病恶化的重要标志.为了更准确地检测出视网膜新生血管,本文提出了一种基于彩色眼底图的视网膜新生血管检测方法.首先通过一种改进的U形卷积神经网络对血管进行分割;然后利用滑动窗口提取特定区域内血管的形态特征,通过支持向量机将窗口内的血管分为普通血管和新生血管.使用来自MESSIDOR数据集和Kaggle数据集的含有视网膜新生血管的彩色眼底图对实验进行训练和测试,结果表明该方法对视网膜新生血管检测的准确率为95.96%;该方法在糖网病计算机辅助诊断方面有潜在的应用前景.Diabetic retinopathy(DR)is one of the major causes of blindness,and the appearance of retinal neovascularization(RN)is an important sign of DR deterioration.In order to detect RN more accurately,a method based on color fundus photograph for retinal neovascularization detection is proposed.First,an improved U-shaped convolutional neural network is used to segment the blood vessels.Then,a sliding window is used to extract the morphological characteristics of blood vessels in the specific area.A support vector machine(SVM)is used to classify the blood vessels into normal vessels and retinal neovascularization in the window.The experiments use color fundus photographs with retinal neovascularization from the MESSIDOR dataset and the Kaggle dataset for training and testing.The result shows that the accuracy of this method for the RN detection is 95.96%;This method has potential application prospects in the computer-aided diagnosis of diabetic retinopathy.

关 键 词：视网膜新生血管检测 血管分割 U形网络 深度学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]