融合注意力的ConvNeXt视网膜病变自动分级  

Automatic classification of retinopathy with attention ConvNeXt

作　　者：黄文博[1] 黄钰翔 姚远[2] 燕杨[1] HUANG Wenbo;HUANG Yuxiang;YAO Yuan;YAN Yang(School of Computer Science and Technology,Changchun Normal University,Changchun 130032,China;Bureau of Major Tasks,Chinese Academy of Sciences,Beijing 100864,China)

机构地区：[1]长春师范大学计算机科学与技术学院,吉林长春130032 [2]中国科学院重大任务局,北京100864

出　　处：《光学精密工程》2022年第17期2147-2154,共8页Optics and Precision Engineering

基　　金：吉林省自然科学基金资助项目(No.YDZJ202101ZYTS147);吉林省教育厅科学研究规划项目(No.JJKH20200830KJ);吉林省教育厅科学技术研究项目(No.JJKH20210887KJ)。

摘　　要：由于视网膜病变的类间图像特征差别小及分类临界值相对模糊,自动分级算法存在识别与分级准确率低的问题,提出了融合高效通道注意力(Efficient Channel Attention,ECA)特征的ConvNeXt视网膜病变自动分级模型。针对数据集中数据不足的问题,采用水平翻转左右变换的方法扩充数据,并引入相关数据集来均衡数据的分布。针对眼底图像中出现的图像模糊、光照不均等问题,采用Graham方法对图像进行预处理突出病变特征。提出了融合注意力的ConvNeXt网络来辅助医生诊断视网膜病变,引入ECA机制,并设计了E-Block模块,该模块具有高性能、低参数的特性,能够在训练过程中有效捕捉跨通道交互的信息,同时避免降维。采用迁移学习方法训练网络的所有层参数,加入dropout方法避免ConvNeXt网络的学习能力过强导致的过拟合问题。实验结果表明,所提出的模型敏感性为95.20%,特异度为98.80%,准确率为95.21%。与常用的网络相比,本文方法针对视网膜病变自动分级各项性能指标均有提高。Due to the small differences in image features between classes and the relative fuzzy classification threshold of retinopathy,automatic classification algorithms are challenged by problems related to low recognition and classification accuracy. This paper proposes an automatic classification model for retinopathy based on an improved ConvNeXt network. Aiming at solving the problem of insufficient data in the data set,the horizontal flip left and right transformation method is used to expand the data,and related data sets are introduced to balance data distribution. To solve problems related to image blurring and uneven illumination in the fundus image,the Graham method was used to predict the image. The characteristics of the lesions are also highlighted. In this paper,an attention-fused ConvNeXt network was proposed to assist doctors in diagnosing retinopathy,an efficient channel attention mechanism was introduced,and an EBlock module was designed to channel interaction information while avoiding dimensionality reduction.The transfer learning method was used to train all layer parameters of the network,and the dropout method was added to avoid the overfitting problem caused by the strong learning ability of the ConvNeXt network. The results show that the sensitivity,specificity,and accuracy of the proposed model are 95. 20%,98. 80%,and 95. 21%,respectively. Compared with the ConvNeXt and other networks,the performance indexes of this network model for automatic classification of retinopathy.

关 键 词：视网膜病变识别与分级 迁移学习 ConvNeXt网络 高效通道注意力 E-Block 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]