July 2019
Volume 60, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2019
Prediction of areas at risk of developing geographic atrophy in color fundus images using deep learning
Author Affiliations & Notes
  • Bart Liefers
    EyeNED Research Group, Department of Radiology and Nuclear Medicine, Radboudumc, Nijmegen, Netherlands
  • Johanna Maria Colijn
    Ophthalmology & Epidemiology, Erasmus MC, Rotterdam, Netherlands
  • Cristina González-Gonzalo
    EyeNED Research Group, Department of Radiology and Nuclear Medicine, Radboudumc, Nijmegen, Netherlands
  • Akshayaa Vaidyanathan
    EyeNED Research Group, Department of Radiology and Nuclear Medicine, Radboudumc, Nijmegen, Netherlands
  • Harm van Zeeland
    EyeNED Research Group, Department of Radiology and Nuclear Medicine, Radboudumc, Nijmegen, Netherlands
    Ophthalmology, Radboudumc, Nijmegen, Netherlands
  • Paul Mitchell
    Clinical Ophthalmology & Eye Health, University of Sydney, New South Wales, Australia
  • Caroline C W Klaver
    Ophthalmology & Epidemiology, Erasmus MC, Rotterdam, Netherlands
    Ophthalmology, Radboudumc, Nijmegen, Netherlands
  • Clara I Sanchez
    EyeNED Research Group, Department of Radiology and Nuclear Medicine, Radboudumc, Nijmegen, Netherlands
    Ophthalmology, Radboudumc, Nijmegen, Netherlands
  • Footnotes
    Commercial Relationships   Bart Liefers, None; Johanna Colijn, None; Cristina González-Gonzalo, None; Akshayaa Vaidyanathan, None; Harm van Zeeland, None; Paul Mitchell, None; Caroline Klaver, None; Clara Sanchez, None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 1455. doi:https://doi.org/
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Bart Liefers, Johanna Maria Colijn, Cristina González-Gonzalo, Akshayaa Vaidyanathan, Harm van Zeeland, Paul Mitchell, Caroline C W Klaver, Clara I Sanchez; Prediction of areas at risk of developing geographic atrophy in color fundus images using deep learning. Invest. Ophthalmol. Vis. Sci. 2019;60(9):1455. doi: https://doi.org/.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Exact quantification of areas of geographic atrophy (GA) can provide an important anatomical endpoint for treatment trials. The prediction of areas where GA may develop can provide valuable personalized prognosis and help in the development of targeted treatments to prevent progression and further vision loss. In this work, we present a model based on a deep convolutional neural network (CNN) that predicts the areas of GA within 5 years from baseline using color fundus (CF) images.

Methods : Areas of GA were delineated by 4 to 5 experienced graders in consensus in 377 CF images (252 eyes) collected from the Rotterdam Study and the Blue Mountains Eye Study. Graders made use of multimodal and follow up images when available, using our EyeNED annotation workstation. We identified 84 pairs of images (baseline and follow-up) of the same eye that were acquired with an interval of approximately 5 years. Image registration was performed by identifying corresponding landmarks between the images, allowing to project the delineated GA of the follow-up image onto the baseline image.
Next, a fully automatic segmentation model, based on a deep CNN, was developed. The CNN was trained to simultaneously segment the current GA area and the area at risk of developing GA, using only the baseline image as input. A five-fold cross-validation was performed to validate the prediction performance.

Results : The model achieved an average dice coefficient of 0.63 for segmentation of areas at risk of developing GA in the 84 images. The intraclass correlation coefficient between the GA area defined by the consensus grading of the follow-up image and the automatically predicted area based on the baseline image was 0.54.

Conclusions : We present a model based on a deep CNN that is capable of identifying areas where GA may develop from CF images. The proposed approach constitutes a step towards personalized prognosis and possible treatment decisions. Furthermore, the model may be used for automatic discovery of new predictive biomarkers for development and growth rate of GA, and may help to automatically identify individuals at risk of developing GA.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Example baseline images with overlay of annotated GA (center) and automatically segmented GA (right). Green indicates current GA, red indicates GA in follow-up. Dice scores for predicted area at risk of developing GA for these images were 0.80, 0.85 and 0.85 respectively.

Example baseline images with overlay of annotated GA (center) and automatically segmented GA (right). Green indicates current GA, red indicates GA in follow-up. Dice scores for predicted area at risk of developing GA for these images were 0.80, 0.85 and 0.85 respectively.

×
×

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.

×