Investigative Ophthalmology & Visual Science Cover Image for Volume 65, Issue 7
June 2024
Volume 65, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2024
Evaluating Performance of ResNet-50 and RETFound Image Classification Algorithms on Ultra-widefield Fundus Images
Author Affiliations & Notes
  • Nikhil Varma Gadiraju
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • David Kuo
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Vahid Ownagh
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Terry Lee
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Lindsey A Chew
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Grant A. Justin
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Nita Valikodath
    Kellogg Eye Center, University of Michigan, Ann Arbor, Michigan, United States
  • Stephanie Hsu
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Miroslav Pajic
    Department of Electrical and Computer Engineering, Duke University, Durham, North Carolina, United States
  • Majda Hadziahmetovic
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Lejla Vajzovic
    Duke Eye Center, Duke University, Durham, North Carolina, United States
  • Footnotes
    Commercial Relationships   Nikhil Gadiraju None; David Kuo None; Vahid Ownagh None; Terry Lee None; Lindsey Chew None; Grant Justin None; Nita Valikodath None; Stephanie Hsu None; Miroslav Pajic None; Majda Hadziahmetovic None; Lejla Vajzovic None
  • Footnotes
    Support  VitreoRetinal Surgery Foundation Research Awards (to NG and LAC), VitreoRetinal Surgery Foundation Research Fellowship (to VO)
Investigative Ophthalmology & Visual Science June 2024, Vol.65, 1605. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Nikhil Varma Gadiraju, David Kuo, Vahid Ownagh, Terry Lee, Lindsey A Chew, Grant A. Justin, Nita Valikodath, Stephanie Hsu, Miroslav Pajic, Majda Hadziahmetovic, Lejla Vajzovic; Evaluating Performance of ResNet-50 and RETFound Image Classification Algorithms on Ultra-widefield Fundus Images. Invest. Ophthalmol. Vis. Sci. 2024;65(7):1605.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : AI-assisted interpretation of ultra-widefield fundus (UWF) photography can enhance pediatric retinal disease screening, leading to earlier disease detection. RETFound is a large Vision Transformer (ViT-L) model pre-trained on 1.6 million color fundus photographs and OCT B-scans, demonstrating strong performance in many tasks (i.e. grading diabetic retinopathy, detecting ischemic stroke and heart failure). An open question is the generalizability of RETFound to classification tasks based on UWF images that were not part of its pre-training dataset. We hypothesize that RETFound will outperform ImageNet-pretrained (IN-p) ViT-L and ResNet-50 models to identify pediatric retinal pathology from a given set of UWF images due to its retina-specific pretraining.

Methods : 304 UWF images were retrospectively collected from 62 healthy pediatric patients undergoing routine eye examination at DukeHealth from 2014-2020; class labels of normal and abnormal were assigned by 2 independent vitreoretinal fellows. 3 machine learning models (RETFound, IN-p ViT-L, and IN-p ResNet) were fine-tuned on the binary image classification task of identifying normal vs. abnormal fundus images using validation set performance. The model with the highest validation accuracy from each original model was assessed on the held-out test set. Performance metrics included accuracy, area under the receiver operating characteristic curve (AUROC), average precision, F1 score, precision, and recall.

Results : UWF images were randomly divided into training, validation, and test datasets. 82 images (27%) were labeled as normal and 222 images (73%) as abnormal. For ResNet, ViT-L, and RETFound models respectively, we report the following: classification accuracy (66.7%, 64.1%, 81.3%), AUROC (77.8%, 76.4%, 81.5%), F1 scores (61.0%, 61.1%, 77.7%), average precision values (73.7%, 67.1%, and 77.6%), precision scores (61.6%, 60.7%, 75.8%), and recall scores (66.6%, 64.1%, 81.3%).

Conclusions : RETFound outperformed ResNet-50 and ViT-L models across all metrics. Shared architecture between RETFound and ViT-L suggests that retina-specific pretraining, without UWF images, improves classification accuracy compared to conventional IN-p. This study underscores RETFound’s potential as an optimal basis for AI analysis of UWF images.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×