Journal article
Deep Learning Approach Readily Differentiates Papilledema, NAION, and Healthy Eyes
American journal of ophthalmology, Vol.276, pp.99-108
08/2025
DOI: 10.1016/j.ajo.2025.04.006
PMID: 40220884
Abstract
Deep learning (DL) has been used in differentiating a range of ophthalmic conditions. We describe a model to distinguish among fundus photos of acquired optic disc swelling due to idiopathic intracranial hypertension (IIH), non-arteritic anterior ischemic optic neuropathy (NAION), and healthy eyes.
Development and validation of a deep-learning diagnostic algorithm.
Our model was trained and validated on 15,088 fundus photos from 5866 eyes, including eyes with IIH with a Frisén grade ≥ 1 (418), acute NAION (780), and healthy controls (4668). We performed external validation on an additional 1126 photos from 928 eyes across these groups. All images were obtained from randomized and nonrandomized clinical trials, publicly available datasets, and real-world clinical sources.
After preprocessing images to standardize optic disc position, we fine-tuned a ResNet-50 model. Performance was evaluated using 5-fold cross-validation, with metrics such as accuracy, AUC-ROC, precision, recall, F1 scores, and confusion matrices calculated. We applied gradient-weighted class activation mapping (Grad-CAM) to generate visual activation maps highlighting the regions of interest in the fundus images. External validation evaluation was performed with majority voting of our cross-validated models.
The primary outcome measures were the model's overall accuracy, class-wide AUC-ROC, precision, recall, F1 scores, and confusion matrices.
The model achieved an internal validation accuracy of 96.2%, with a macro-average AUC-ROC of 0.995. F1 scores ranged from 0.90 to 0.97 for all classes. The external validation set had an accuracy of 93.6%, F1 scores from 0.90 to 0.95 and macro-average AUC-ROC of 0.980. Activation maps consistently highlighted the optic disc, with emphasis on the inferior optic disc for IIH, superior optic disc for NAION and the entire optic disc for healthy eyes.
Our study demonstrates the potential of fundus-based deep learning models to accurately distinguish among IIH, NAION, and healthy eyes, providing a potentially valuable diagnostic method. With its strong discriminative capabilities, this model can be an important tool for neuro-ophthalmic assessment, particularly when access to specialized neuro-ophthalmologists is limited.
Details
- Title: Subtitle
- Deep Learning Approach Readily Differentiates Papilledema, NAION, and Healthy Eyes
- Creators
- David Szanto - Icahn School of Medicine at Mount SinaiAsala Erekat - University Hospital GalwayBrian WoodsJui-Kai Wang - The University of Texas Southwestern Medical CenterMona GarvinBrett A Johnson - University of IowaRandy KardonEdward LintonMark J Kupersmith - Icahn School of Medicine at Mount Sinai
- Resource Type
- Journal article
- Publication Details
- American journal of ophthalmology, Vol.276, pp.99-108
- DOI
- 10.1016/j.ajo.2025.04.006
- PMID
- 40220884
- NLM abbreviation
- Am J Ophthalmol
- ISSN
- 0002-9394
- eISSN
- 1879-1891
- Publisher
- ELSEVIER SCIENCE INC
- Grant note
- New York Eye and Ear Infirmary Foundation (New York, NY)Shulman Family NAION Fund (New York, NY)Neuro-Ophthalmology Research Fund of New York Eye and Ear Infirmary (New York, NY)Research to Prevent Blindness, Inc. (New York, NY): R01 EY015473 Department of Veterans Affairs Center for the Prevention and Treat-ment of Visual Loss, Rehabilitation Research and Development: I50 RX003002 Veterans Affairs Rehabilitation Research and Development: I01RX003797 National Center for Advancing Translational Sciences of the National Institutes of Health (NIH): NIH CTSA UL1TR002537 Health Research Board: ICAT-2022-001 ICAT ProgramBarry Family Center for Ophthalmic Artificial Intelligence and Human Health (New York, NY)
The New York Eye and Ear Infirmary Foundation (New York, NY) ; Shulman Family NAION Fund (New York, NY) ; Neuro-Ophthalmology Research Fund of New York Eye and Ear Infirmary (New York, NY) ; Research to Prevent Blindness, Inc. (New York, NY) unrestricted grant to the Department of Ophthalmology; National Eye Institute R01 EY015473 ; Department of Veterans Affairs Center for the Prevention and Treat-ment of Visual Loss, Rehabilitation Research and Development I50 RX003002 ; Veterans Affairs Rehabilitation Research and Development I01RX003797 . Research reported in this publication was supported in part by the National Center for Advancing Translational Sciences of the National Institutes of Health (NIH) under Award Number NIH CTSA UL1TR002537 . The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. This research was partially funded by the Health Research Board (ICAT-2022-001) and the ICAT Program. The Barry Family Center for Ophthalmic Artificial Intelligence and Human Health (New York, NY) . Financial Disclosures: The authors indicate no conflicts of interest. All authors attest that they meet the current ICMJE criteria for authorship.
- Language
- English
- Electronic publication date
- 04/10/2025
- Date published
- 08/2025
- Academic Unit
- Electrical and Computer Engineering; The University of Iowa Institute for Vision Research; Iowa Neuroscience Institute; Ophthalmology and Visual Sciences
- Record Identifier
- 9984811225202771
Metrics
19 Record Views