Joint grading of diabetic retinopathy and macular edema based on improved ResNeSt50
10.3969/j.issn.1005-202X.2025.06.009
- VernacularTitle:基于改进ResNeSt50对糖尿病视网膜病变和黄斑水肿的联合分级
- Author:
Yuxuan LIU
1
;
De GU
1
Author Information
1. 江南大学物联网工程学院,江苏 无锡 214122
- Publication Type:Journal Article
- Keywords:
diabetes;
retina;
ResNeSt50;
partial convolution;
attention mechanism
- From:
Chinese Journal of Medical Physics
2025;42(6):766-774
- CountryChina
- Language:Chinese
-
Abstract:
To address the challenge of joint grading caused by the diverse lesion morphologies associated with different stages of diabetic retinopathy,such as hemorrhages,microaneurysms,and neovascularization,which often obscure the lesions of diabetic macular edema,an improved ResNeSt50-based joint grading network is proposed.The modified ResNeSt50 with a novel convolutional operation(partial convolution)replacing the standard 3×3 convolution in the residual blocks is used to extract image features for exploring the specificities of diabetic retinopathy and macular edema.Subsequently,a disease-related attention module is introduced to capture the intrinsic correlation between diabetic retinopathy and macular edema.Experiments conducted on the Messidor and IDRID datasets show that the proposed approach achieves joint accuracies of 83.7%and 64.1%,respectively.The results demonstrate that the proposed algorithm can significantly improve the performance of joint grading for diabetic retinopathy and diabetic macular edema.
