Your new experience awaits. Try the new design now and help us make it even better

ORIGINAL RESEARCH article

Front. Med.

Sec. Ophthalmology

Volume 12 - 2025 | doi: 10.3389/fmed.2025.1584378

This article is part of the Research TopicEfficient Artificial Intelligence in Ophthalmic Imaging – Volume IIView all 6 articles

Development and Evaluation of a Deep Learning System for Screening Real-world Multiple Abnormal Findings Based on Ultra-Widefield Fundus Images

Provisionally accepted
  • 1Department of Ophthalmology, Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
  • 2Beijing Airdoc Technology Co., Ltd., Beijing, China
  • 3Monash Medical AI Group, Monash University, Clayton, Australia
  • 4Department of Ophthalmology, Shanghai Children’s Hospital, School of medicine, Shanghai Jiao Tong University, Shanghai, China
  • 5Zhenjiang Ruikang Hospital, Zhenjiang, China
  • 6Shibei Hospital Jingan District Shanghai, Shanghai, China
  • 7Medical College of China Three Gorges University, Yichang, China

The final, formatted version of the article will be published soon.

Purpose: To develop and evaluate a deep learning system for screening multiple abnormal findings including hemorrhages, drusen, hard exudates, cotton wool spots and retinal breaks using ultra-widefield fundus images.Methods: The system consisted of three modules: (I) quality assessment module, (II) artifact removal module and (III) lesion recognition module. In Module III, a heatmap was generated to highlight the lesion area. A total of 4521 UWF images were used for the training and internal validation of the DL system. The system was evaluated in 2 external validation datasets consisting of 344 images and 894 images from two other hospitals. The performance of the system in these 2 datasets was compared with or without Module II.Results: In both external validation datasets, the deep learning system made better performance when recognizing lesions on processed images after Module II than on original images without Module II. Module II-enhanced preprocessing improved Module III's five-lesion recognition performance by an average of 6.73% and 14.4% areas under the curves, 14.47% and 19.62% accuracy in the two external validations.Conclusions: Our system showed reliable performance for detecting MAF in real-world UWF images. For deep learning systems to recognize real-world images, the artifact removal module was indeed helpful.

Keywords: deep learning, ultra-widefield fundus images, Real-world, artifact removal, Multiple abnormal findings

Received: 27 Feb 2025; Accepted: 05 May 2025.

Copyright: © 2025 Xiao, Ju, Lu, Zhang, Jiang, Yang, Zhang, Zhang, Liu, Liang, Ren, Yin, Liu, Tong, Wang, Feng, Song, Chen, Ge, Shao, Peng, Chen and Zhao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence:
Jie Peng, Department of Ophthalmology, Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
Jili Chen, Shibei Hospital Jingan District Shanghai, Shanghai, China
Peiquan Zhao, Department of Ophthalmology, Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.