Current studentsExisting staffContact us
 

Applying supervised contrastive learning for the detection of diabetic retinopathy and its severity levels from fundus images

Simple item page
dc.contributor.authorIslam, M.R.
dc.contributor.authorAbdulrazak, L.F.
dc.contributor.authorNahiduzzaman, M.
dc.contributor.authorGoni, M.O.F.
dc.contributor.authorAnower, M.S.
dc.contributor.authorAhsan, M.
dc.contributor.authorHaider, J.
dc.contributor.authorKowalski, M.
dc.date.issued2022
dc.description.abstractDiabetic Retinopathy (DR) is a major complication in human eyes among the diabetic patients. Early detection of the DR can save many patients from permanent blindness. Various artificial intelligent based systems have been proposed and they outperform human analysis in accurate detection of the DR. In most of the traditional deep learning models, the cross-entropy is used as a common loss function in a single stage end-to-end training method. However, it has been recently identified that this loss function has some limitations such as poor margin leading to false results, sensitive to noisy data and hyperparameter variations. To overcome these issues, supervised contrastive learning (SCL) has been introduced. In this study, SCL method, a two-stage training method with supervised contrastive loss function was proposed for the first time to the best of authors' knowledge to identify the DR and its severity stages from fundus images (FIs) using “APTOS 2019 Blindness Detection” dataset. “Messidor-2” dataset was also used to conduct experiments for further validating the model's performance. Contrast Limited Adaptive Histogram Equalization (CLAHE) was applied for enhancing the image quality and the pre-trained Xception CNN model was deployed as the encoder with transfer learning. To interpret the SCL of the model, t-SNE method was used to visualize the embedding space (unit hyper sphere) composed of 128 D space into a 2 D space. The proposed model achieved a test accuracy of 98.36%, and AUC score of 98.50% to identify the DR (Binary classification) and a test accuracy of 84.364%, and AUC score of 93.819% for five stages grading with the APTOS 2019 dataset. Other evaluation metrics (precision, recall, F1-score) were also determined with APTOS 2019 as well as with Messidor-2 for analyzing the performance of the proposed model. It was also concluded that the proposed method achieved better performance in detecting the DR compared to the conventional CNN without SCL and other state-of-the-art methods.
dc.identifier.citationComputers in Biology and Medicine, 2022; 146(article no. 105602):1-18
dc.identifier.doi10.1016/j.compbiomed.2022.105602
dc.identifier.issn0010-4825
dc.identifier.issn1879-0534
dc.identifier.urihttps://hdl.handle.net/11541.2/41928
dc.language.isoen
dc.publisherElsevier
dc.rightsCopyright 2022 The author(s) (https://creativecommons.org/licenses/by-nc-nd/4.0/) Access Condition Notes: This is an open access article
dc.source.urihttps://doi.org/10.1016/j.compbiomed.2022.105602
dc.subjectdiabetic retinopathy
dc.subjectimage analysis
dc.subjectcnn
dc.subjectsupervised contrastive learning
dc.subjectclahe
dc.subjectt-SNE
dc.titleApplying supervised contrastive learning for the detection of diabetic retinopathy and its severity levels from fundus images
dc.typeJournal article
pubs.publication-statusPublished
ror.mmsid9916946119001831

Files

Collections