June 2017
Volume 58, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2017
Deep learning is effective in classifying normal versus Age-related Macular degeneration using OCT images
Author Affiliations & Notes
  • Aaron Y Lee
    University of Washington, Seattle, Washington, United States
  • Footnotes
    Commercial Relationships   Aaron Lee, Novartis (F)
  • Footnotes
    Support  Unrestricted grant from RPB to University of Washington
Investigative Ophthalmology & Visual Science June 2017, Vol.58, 822. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Aaron Y Lee; Deep learning is effective in classifying normal versus Age-related Macular degeneration using OCT images. Invest. Ophthalmol. Vis. Sci. 2017;58(8):822.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : To determine if deep learning could be utilized to distinguish normal OCT images from images of patients with Age-related Macular Degeneration (AMD).

Methods : Automated extraction of an OCT imaging database was performed and linked to clinical endpoints from the EMR. OCT macular volumes were obtained by Heidelberg Spectralis, and each volume was linked to EMR clinical endpoints extracted from EPIC. The central 11 images were selected from each OCT volume of two cohorts of patients: normal and AMD. Cross-validation was performed using a random subset of patients. Receiver operator curves (ROC) were constructed at an independent image level, macular OCT volume level, and patient level.

Results : Of a recent extraction of 2.6 million OCT images linked to clinical datapoints from the EMR, 52,690 normal macular OCT images and 48,312 AMD macular OCT images were selected. A deep neural network was trained to categorize images as either normal or AMD. At the image level, we achieved an area under the ROC of 92.78% with an accuracy of 87.63%. At the OCT volume level, we achieved an area under the ROC of 93.83% with an accuracy of 88.98%. At a patient level, we achieved an area under the ROC of 97.45% with an accuracy of 93.45%. Peak sensitivity and specificity with optimal cutoffs were 92.64% and 93.69% respectively.

Conclusions : Deep learning techniques achieve high accuracy and is effective as a new image classification technique. These findings have important implications in utilizing OCT in automated screening and the development of computer aided diagnosis tools in the future.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.

 

Figure 1. Learning curve by cross validation (A). Loss function over iterations (B).

Figure 1. Learning curve by cross validation (A). Loss function over iterations (B).

 

Figure 2. OCT images (A, B, C) with Age-Related Macular Degeneration. After occlusion based testing of deep learning algorithm a overlying heatmap of probabilities (D, E, F) were generated showing where deep learning was most dependent on to make the diagnosis of AMD.

Figure 2. OCT images (A, B, C) with Age-Related Macular Degeneration. After occlusion based testing of deep learning algorithm a overlying heatmap of probabilities (D, E, F) were generated showing where deep learning was most dependent on to make the diagnosis of AMD.

×
×

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.

×