Using Deep Neural Networks to Classify the Severity of Diabetic Retinopathy

Chun Ying Chen, Meng Chou Chang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

Diabetic Retinopathy (DR) is a type of retinopathy resulted from diabetic hyperglycemia (high blood glucose), which causes the abnormal blood vessels of eyes to bleed and causes the patients to lose sight. In this paper, we employ two deep neural networks, InceptionV3 and EfficientNet, to classify the retinal fundus images of DR patients into five severity levels of DR, and improve the accuracy of the neural networks by techniques such as dropout, data preprocessing, data augmentation, and learning rate adjustment. The experimental results showed that EfficientNetB0 has the best performance with an accuracy of 86.26% and a quadratic weighted kappa of 0.926.

Original languageEnglish
Title of host publicationProceedings - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages241-242
Number of pages2
ISBN (Electronic)9781665470506
DOIs
StatePublished - 2022
Externally publishedYes
Event2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022 - Taipei, Taiwan
Duration: 06 07 202208 07 2022

Publication series

NameProceedings - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022

Conference

Conference2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
Country/TerritoryTaiwan
CityTaipei
Period06/07/2208/07/22

Bibliographical note

Publisher Copyright:
© 2022 IEEE.

Keywords

  • Deep neural network
  • deep learning
  • diabetic retinopathy

Fingerprint

Dive into the research topics of 'Using Deep Neural Networks to Classify the Severity of Diabetic Retinopathy'. Together they form a unique fingerprint.

Cite this