loading page

Applying novel self-supervised learning for early detection of retinopathy of prematurity
  • +1
  • Dongmei Wang,
  • Wanli Qiao,
  • Wei Guo,
  • Yuansong Cai
Dongmei Wang
Changchun University of Technology

Corresponding Author:[email protected]

Author Profile
Wanli Qiao
Changchun University of Technology
Author Profile
Wei Guo
Changchun University of Technology
Author Profile
Yuansong Cai
Changchun University of Technology
Author Profile

Abstract

Retinopathy of Prematurity (ROP) mainly occurs in premature infants with low birth weight, and it is the leading cause of childhood blindness. Early and accurate ROP diagnosis is imperative for appropriate treatment. However, less research concentrates on early-stage ROP diagnosis based on limited-labeled images in an imbalanced dataset. To address the dilemma, this study proposed a novel self-supervised network, MOCO-MIM, for early ROP grading. The proposed classification network was evaluated on a total of 553 labeled fundus images from 89 preterm infants. The trained network achieved a test accuracy of 98.29% and an AUC score of 97.6% for three stages of grading. The adopted method is verified that the proposed method can be detected early stages of ROP more efficiently and grade the severity more accurately based on limited-labeled fundus images, which is superior to the existing state-of-the-art methods.
09 May 2024Submitted to Electronics Letters
10 May 2024Submission Checks Completed
10 May 2024Assigned to Editor
29 May 2024Review(s) Completed, Editorial Evaluation Pending
03 Jun 2024Editorial Decision: Revise Minor
09 Jun 20241st Revision Received
13 Jun 2024Submission Checks Completed
13 Jun 2024Assigned to Editor
13 Jun 2024Review(s) Completed, Editorial Evaluation Pending
19 Jun 2024Editorial Decision: Accept