Diagnosing pathologic myopia by identifying morphologic patterns using ultra widefield images with deep learning
Pathologic myopia is a leading cause of visual impairment and blindness. While deep learning-based approaches aid in recognizing pathologic myopia using color fundus photography, they often rely on implicit patterns that lack clinical interpretability. This study aims to diagnose pathologic myopia by identifying clinically significant morphologic patterns, specifically posterior staphyloma and myopic maculopathy, by leveraging ultra-widefield (UWF) images that provide a broad retinal field of view. We curate a large-scale, multi-source UWF myopia dataset called PSMM and introduce RealMNet, an end-to-end lightweight framework designed to identify these challenging patterns. Benefiting from the fast pretraining distillation backbone, RealMNet comprises only 21 million parameters, which facilitates deployment for medical devices. Extensive experiments conducted across three different protocols demonstrate the robustness and generalizability of RealMNet. RealMNet achieves an F1 Score of 0.7970 (95% CI 0.7612-0.8328), mAP of 0.8497 (95% CI 0.8058-0.8937), and AUROC of 0.9745 (95% CI 0.9690-0.9801), showcasing promise in clinical applications.
Authors
- Yingting Zhu (ORCID: https://orcid.org/0009-0002-2783-2889)
- Chengming Yang
- Peiwu Qin (ORCID: https://orcid.org/0000-0002-7829-8973)
- Yehong Zhuo (ORCID: https://orcid.org/0000-0003-3247-7199)
- Chunsheng Qu (ORCID: https://orcid.org/0000-0002-7734-9302)
- Z. Qian (ORCID: https://orcid.org/0009-0008-1409-9853)
- Zhenqiang Huang
- Zhicheng Du (ORCID: https://orcid.org/0009-0002-5318-6809)
- Kankan Zhao (ORCID: https://orcid.org/0000-0003-2116-017X)
- Shuai Wang (ORCID: https://orcid.org/0000-0003-3730-6401)
- Chenggang Clarence Yan (ORCID: https://orcid.org/0000-0003-1204-0512)
- Dongmei Yu (ORCID: https://orcid.org/0000-0001-7901-4365)
- Yang Liu (ORCID: https://orcid.org/0000-0002-3087-242X)
- Lihui Luo (ORCID: https://orcid.org/0000-0002-9471-7373)
- Xingru Huang (ORCID: https://orcid.org/0000-0003-3971-8434)
- Duanpo Wu (ORCID: https://orcid.org/0000-0001-6954-6587)
- Chenying Lu
- Jiaqi Chen (ORCID: https://orcid.org/0000-0003-3134-2436)
- Ziheng Zhang (ORCID: https://orcid.org/0000-0002-5938-4729)
- Jiantao Wang (ORCID: https://orcid.org/0009-0000-6071-0969)
- Minjiang Chen (ORCID: https://orcid.org/0000-0002-0814-9139)
- Cuiping Jiang (ORCID: https://orcid.org/0000-0001-9694-5399)
- Simin Deng (ORCID: https://orcid.org/0009-0000-0663-7078)
- Jiansong Ji (ORCID: https://orcid.org/0000-0003-4727-2794)
Institutions
- Shandong University (CN)
- Sun Yat-sen University (CN)
- Jinan University (CN)
- Southern University of Science and Technology (CN)
- Lishui Central Hospital (CN)
- Tsinghua–Berkeley Shenzhen Institute (CN)
- Hangzhou Dianzi University (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- npj Digital Medicine
- Published
- 2025-07-13
- DOI
- https://doi.org/10.1038/s41746-025-01849-y
- Citations
- 5
- Primary Topic
- Retinal Imaging and Analysis
- Type
- article
- Field-Weighted Citation Impact
- 6.80
Funders
- National Natural Science Foundation of China
- Science, Technology and Innovation Commission of Shenzhen Municipality
- Tsinghua Shenzhen International Graduate School