计算机科学

计算机科学 ›› 2026, Vol. 53 ›› Issue (6A): 250200048-6.doi: 10.11896/jsjkx.250200048

• 图像处理&多媒体技术 • 上一篇    下一篇

基于标签松弛多视角特征融合的糖尿病视网膜病变分级

段练   

  1. 南通大学医学信息学系 江苏 南通 214122
  • 出版日期:2026-06-16 发布日期:2026-06-12
  • 通讯作者: 段练(duanlian@ntu.edu.cn)
  • 基金资助:
    江苏高校哲学社会科学研究项目(2023SJYB1680)

Diabetic Retinopathy Grading Based on Label Relaxation Multi-view Feature Fusion

DUAN Lian   

  1. Department of Medical Informatics,Nantong University,Nantong,Jiangsu 214122,China
  • Published:2026-06-16 Online:2026-06-12
  • About author:DUAN Lian,born in 1988,Ph.D,asso-ciate professor.His main research in-terests include machine learning and medical informatics.
  • Supported by:
    Philosophy and Social Sciences Research Project in Higher Education of Jiangsu Province(2023SJYB1680).

PDF (PC)

41

摘要： 糖尿病视网膜病变是糖尿病较为常见的并发症,准确识别糖尿病视网膜病变等级对后续治疗非常关键。眼底图像在糖尿病视网膜病变分级中发挥着关键作用。随着人工智能技术的发展,许多研究者已经从眼底图像中提取深度特征和放射组学特征开展糖尿病视网膜病变分级研究。结合深度特征和放射组学特征,设计了一种特征融合算法。首先,利用卷积神经网络从眼底图像中提取深度特征,并使用放射组学方法提取放射组学特征。随后,设计了一种基于标签松弛的多视角学习算法进行特征融合。标签松弛的主要目标是增强标签空间中训练样本的可区分性,从而提高模型的分类准确率。此外,还引入了基于流形学习方法的图约束,以减轻标签松弛所导致的过拟合问题。最后,在DR1和MESSIDOR两个眼底图像数据集上验证了所提出方法的有效性。

关键词: 糖尿病视网膜病变分级, 多视角特征融合, 放射组学, 深度特征, 标签松弛

Abstract: Diabetic retinopathy is a common complication of diabetes,and accurately identifying the stages of diabetic retinopathy is crucial for subsequent treatment.Fundus images play a key role in the grading of diabetic retinopathy.With the advancement of artificial intelligence technologies,many researchers have extracted deep features and radiomic features from fundus images to conduct studies on the grading of diabetic retinopathy.This study combines deep features and radiomic features to design a feature fusion algorithm.Firstly,deep features are extracted from fundus images using convolutional neural networks,while radiomic features are obtained through radiomic methods.Subsequently,a label relaxation-based multi-view learning algorithm is designed for feature fusion.The primary goal of label relaxation is to enhance the distinguishability of training samples in the label space,thereby improving the classification accuracy of the model.Furthermore,this study introduces a graph constraint based on manifold learning methods to mitigate the overfitting issues caused by label relaxation.Finally,theeffectiveness of the proposed methodis validated on two fundus image datasets:the DR1 dataset and the MESSIDOR dataset.

Key words: Diabetic retinopathy grading, Multi-view feature fusion, Radiomics, Deep features, Label relaxation

中图分类号: 

  • TP391
[1] ABEL E D,GLOYN A L,EVANS-MOLINA C,et al.Diabetesmellitus－Progress and opportunities in the evolving epidemic[J].Cell,2024,187(15):3789-3820.
[2] YAMENY A A.Diabetes Mellitus Overview 2024[J].Journal ofBioscience and Applied Research,2024,10(3):641-645.
[3] MUTHUSAMY D,PALANI P.Deep learning model using classification for diabetic retinopathy detection:an overview[J].Artificial Intelligence Review,2024,57(7):185.
[4] DAI L,SHENG B,CHEN T,et al.A deep learning system for predicting time to progression of diabetic retinopathy[J].Nature Medicine,2024,30(2):584-594.
[5] BHATI A,GOUR N,KHANNA P,et al.An interpretable dual attention network for diabetic retinopathy grading:IDANet[J].Artificial Intelligence in Medicine,2024,149:102782.
[6] NAYAK J,BHAT P S,ACHARYA U R,et al.Automatedidentification of diabetic retinopathy stages using digital fundus images[J].Journal of Medical Systems,2008,32:107-115.
[7] DE LA CALLEJA J,TECUAPETLA L,AUXILIO MEDINAM,et al.LBP and machine learning for diabetic retinopathy detection[C]//Intelligent Data Engineering and Automated Learning-IDEAL 2014:15th International Conference.Springer International Publishing,2014:110-117.
[8] TING D S W,CHEUNG C Y L,LIM G,et al.Development and validation of a deep learning system for diabetic retinopathy and related eye diseases using retinal images from multiethnic populations with diabetes[J].Jama,2017,318(22):2211-2223.
[9] KRAUSE J,GULSHAN V,RAHIMY E,et al.Grader variability and the importance of reference standards for evaluating machine learning models for diabetic retinopathy[J].Ophthalmology,2018,125(8):1264-1272.
[10] SUGENO A,ISHIKAWA Y,OHSHIMA T,et al.Simple methods for the lesion detection and severity grading of diabetic retinopathy by image processing and transfer learning[J].Compu-ters in Biology and Medicine,2021,137:104795.
[11] BELLEMO V,LIM Z W,LIM G,et al.Artificial intelligenceusing deep learning to screen for referable and vision-threatening diabetic retinopathy in Africa:a clinical validation study[J].The Lancet Digital Health,2019,1(1):e35-e44.
[12] TARIQ H,RASHID M,JAVED A,et al.Performance analysis of deep-neural-network-based automatic diagnosis of diabetic retinopathy[J].Sensors,2021,22(1):205.
[13] ROCHA A,CARVALHO T,JELINEK H F,et al.Points of interest and visual dictionaries for automatic retinal lesion detection[J].IEEE Transactions on Biomedical Engineering,2012,59(8):2244-2253.
[14] DECENCIÈRE E,ZHANG X,CAZUGUEL G,et al.Feedback on a publicly distributed image database:the Messidor database[J].Image Analysis & Stereology,2014,33(3):231-234.
[15] LIANG X,ALSHEMMARY E N,MA M,et al.Automatic Diabetic Foot Prediction Through Fundus Images by Radiomics Features[J].IEEE Access,2021,9:92776-92787.
[16] ZIMMERMAN J B,COUSINS S B,HARTZELL K M,et al.A psychophysical comparison of two methods for adaptive histogram equalization[J].Journal of Digital Imaging,1989,2:82-91.
[17] FANG X,XU Y,LI X,et al.Regularized label relaxation linear regression[J].IEEE Transactions on Neural Networks and Learning Systems,2017,29(4):1006-1018.
[18] XIE M,HAN Z,ZHANG C,et al.Exploring and exploiting uncertainty for incomplete multi-view classification[C]//Procee-dings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2023:19873-19882.
[19] ZHOU H,XUE Z,LIU Y,et al.RTMC:A Rubost TrustedMulti-View Classification Framework[C]//2023 IEEE International Conference on Multimedia and Expo(ICME).IEEE,2023:576-581.
[20] LUO Y,HUANG Q,LIU L.Classification of tumor in one single ultrasound image via a novel multi-view learning strategy[J].Pattern Recognition,2023,143:109776.
[21] XU C,SI J,GUAN Z,et al.Reliable conflictive multi-view lear-ning[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2024:16129-16137.
[22] ZHANG C,CUI Y,HAN Z,et al.Deep partial multi-view learning[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,44(5):2402-2415.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市两江新区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发