Display options
Share it on

Front Neuroinform. 2021 Dec 24;15:778552. doi: 10.3389/fninf.2021.778552. eCollection 2021.

DR-IIXRN : Detection Algorithm of Diabetic Retinopathy Based on Deep Ensemble Learning and Attention Mechanism.

Frontiers in neuroinformatics

Zhuang Ai, Xuan Huang, Yuan Fan, Jing Feng, Fanxin Zeng, Yaping Lu

Affiliations

  1. Department of Research and Development, Sinopharm Genomics Technology Co., Ltd., Jiangsu, China.
  2. Department of Ophthalmology, Beijing Chao-Yang Hospital, Capital Medical University, Beijing, China.
  3. Medical Research Center, Beijing Chao-Yang Hospital, Capital Medical University, Beijing, China.
  4. Department of Clinical Research Center, Dazhou Central Hospital, Sichuan, China.

PMID: 35002666 PMCID: PMC8740273 DOI: 10.3389/fninf.2021.778552

Abstract

Diabetic retinopathy (DR) is one of the common chronic complications of diabetes and the most common blinding eye disease. If not treated in time, it might lead to visual impairment and even blindness in severe cases. Therefore, this article proposes an algorithm for detecting diabetic retinopathy based on deep ensemble learning and attention mechanism. First, image samples were preprocessed and enhanced to obtain high quality image data. Second, in order to improve the adaptability and accuracy of the detection algorithm, we constructed a holistic detection model DR-IIXRN, which consists of Inception V3, InceptionResNet V2, Xception, ResNeXt101, and NASNetLarge. For each base classifier, we modified the network model using transfer learning, fine-tuning, and attention mechanisms to improve its ability to detect DR. Finally, a weighted voting algorithm was used to determine which category (normal, mild, moderate, severe, or proliferative DR) the images belonged to. We also tuned the trained network model on the hospital data, and the real test samples in the hospital also confirmed the advantages of the algorithm in the detection of the diabetic retina. Experiments show that compared with the traditional single network model detection algorithm, the auc, accuracy, and recall rate of the proposed method are improved to 95, 92, and 92%, respectively, which proves the adaptability and correctness of the proposed method.

Copyright © 2021 Ai, Huang, Fan, Feng, Zeng and Lu.

Keywords: attention mechanism; deep learning; diabetic retinopathy; ensemble learning; image processing

Conflict of interest statement

ZA, YF, and YL are employees of Sinopharm Genomics Technology Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that c

References

  1. IEEE J Biomed Health Inform. 2014 Sep;18(5):1717-28 - PubMed
  2. IEEE J Biomed Health Inform. 2021 Aug;25(8):2988-2996 - PubMed
  3. IEEE Trans Med Imaging. 2018 Nov;37(11):2493-2501 - PubMed
  4. Med Image Anal. 2021 Aug;72:102118 - PubMed
  5. IEEE Trans Neural Netw Learn Syst. 2019 Feb;30(2):553-565 - PubMed
  6. Plant Methods. 2021 Apr 1;17(1):36 - PubMed
  7. IEEE J Biomed Health Inform. 2017 Jan;21(1):193-200 - PubMed
  8. BMJ. 2020 Apr 28;369:m997 - PubMed
  9. Comput Biol Med. 2020 Jun;121:103795 - PubMed
  10. Med Image Anal. 2020 Jan;59:101561 - PubMed
  11. Med Image Anal. 2020 Jul;63:101715 - PubMed
  12. Pattern Anal Appl. 2021 May 9;:1-14 - PubMed
  13. IEEE J Biomed Health Inform. 2021 Jan;25(1):77-87 - PubMed
  14. Cell. 2018 Feb 22;172(5):1122-1131.e9 - PubMed

Publication Types