Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading
Diabetic Retinopathy (DR) causes quite a few blindness worldwide, which can be refrained by the timely diagnosis on retinal images. Recently, researches on deep learning-based retinal image classification have accelerated outstanding improvements in DR grading task. However, existing DR grading work...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9567704/ |
| _version_ | 1819034516233125888 |
|---|---|
| author | Yujie Li Zhang Song Sunkyoung Kang Sungtae Jung Wenpei Kang |
| author_facet | Yujie Li Zhang Song Sunkyoung Kang Sungtae Jung Wenpei Kang |
| author_sort | Yujie Li |
| collection | DOAJ |
| description | Diabetic Retinopathy (DR) causes quite a few blindness worldwide, which can be refrained by the timely diagnosis on retinal images. Recently, researches on deep learning-based retinal image classification have accelerated outstanding improvements in DR grading task. However, existing DR grading works are mostly limited to a supervised manner. They require accurately annotated data labeled by professional experts, and the annotating work is very laborious and time-consuming. We propose a Semi-supervised Auto-encoder Graph Network (SAGN) for the challenging DR diagnosis to relax this constraint. Precisely, SAGN consists of three major modules: auto-encoder feature learning, neighbor correlation mining, and graph representation. Firstly, our model learns to extract representations from retinal images and reconstruct them as close to original inputs as possible. Then neighbor correlations among labeled and unlabeled samples are established by their similarities, calculated by the radial basis function. Finally, we operate Graph Convolutional Neural Network (GCN) to grade retinal samples from extracted features and their correlations. To evaluate the performance of SAGN, we conduct sufficient comparative experiments on APTOS 2019 dataset, trained from EyePACS. Results demonstrate that our SAGN model can achieve comparable performance with limited labeled retinal images with the help of large amounts of unlabeled data. |
| first_indexed | 2024-12-21T07:34:58Z |
| format | Article |
| id | doaj.art-3282de729f1d40ea9d66c38c29d687e0 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-21T07:34:58Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-3282de729f1d40ea9d66c38c29d687e02022-12-21T19:11:28ZengIEEEIEEE Access2169-35362021-01-01914075914076710.1109/ACCESS.2021.31194349567704Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy GradingYujie Li0https://orcid.org/0000-0001-8195-4655Zhang Song1https://orcid.org/0000-0002-8240-1742Sunkyoung Kang2https://orcid.org/0000-0002-5888-3710Sungtae Jung3Wenpei Kang4https://orcid.org/0000-0002-0523-4960Weifang Key Laboratory of Blockchain on Agricultural Vegetables, Weifang University of Science and Technology, Weifang, Shandong, ChinaThe Affiliated Hospital of Qingdao University, Qingdao, ChinaDepartment of Computer Software Engineering, Wonkwang University, Iksan, Jeonbuk, Republic of KoreaDepartment of Computer Software Engineering, Wonkwang University, Iksan, Jeonbuk, Republic of KoreaBusiness College of Southwest University, Chongqing, ChinaDiabetic Retinopathy (DR) causes quite a few blindness worldwide, which can be refrained by the timely diagnosis on retinal images. Recently, researches on deep learning-based retinal image classification have accelerated outstanding improvements in DR grading task. However, existing DR grading works are mostly limited to a supervised manner. They require accurately annotated data labeled by professional experts, and the annotating work is very laborious and time-consuming. We propose a Semi-supervised Auto-encoder Graph Network (SAGN) for the challenging DR diagnosis to relax this constraint. Precisely, SAGN consists of three major modules: auto-encoder feature learning, neighbor correlation mining, and graph representation. Firstly, our model learns to extract representations from retinal images and reconstruct them as close to original inputs as possible. Then neighbor correlations among labeled and unlabeled samples are established by their similarities, calculated by the radial basis function. Finally, we operate Graph Convolutional Neural Network (GCN) to grade retinal samples from extracted features and their correlations. To evaluate the performance of SAGN, we conduct sufficient comparative experiments on APTOS 2019 dataset, trained from EyePACS. Results demonstrate that our SAGN model can achieve comparable performance with limited labeled retinal images with the help of large amounts of unlabeled data.https://ieeexplore.ieee.org/document/9567704/Diabetic retinopathy gradingsemi-supervised learningauto-encodergraph convolutional network |
| spellingShingle | Yujie Li Zhang Song Sunkyoung Kang Sungtae Jung Wenpei Kang Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading IEEE Access Diabetic retinopathy grading semi-supervised learning auto-encoder graph convolutional network |
| title | Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading |
| title_full | Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading |
| title_fullStr | Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading |
| title_full_unstemmed | Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading |
| title_short | Semi-Supervised Auto-Encoder Graph Network for Diabetic Retinopathy Grading |
| title_sort | semi supervised auto encoder graph network for diabetic retinopathy grading |
| topic | Diabetic retinopathy grading semi-supervised learning auto-encoder graph convolutional network |
| url | https://ieeexplore.ieee.org/document/9567704/ |
| work_keys_str_mv | AT yujieli semisupervisedautoencodergraphnetworkfordiabeticretinopathygrading AT zhangsong semisupervisedautoencodergraphnetworkfordiabeticretinopathygrading AT sunkyoungkang semisupervisedautoencodergraphnetworkfordiabeticretinopathygrading AT sungtaejung semisupervisedautoencodergraphnetworkfordiabeticretinopathygrading AT wenpeikang semisupervisedautoencodergraphnetworkfordiabeticretinopathygrading |