July 2018
Volume 59, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2018
Artificial intelligence in retinopathy of prematurity: development of a fully automated deep convolutional neural network (DeepROP) for plus disease diagnosis
Author Affiliations & Notes
  • James M Brown
    Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, Massachusetts, United States
  • J. Peter Campbell
    Ophthalmology, Oregon Health & Science University, Portland, Oregon, United States
  • Susan Ostmo
    Ophthalmology, Oregon Health & Science University, Portland, Oregon, United States
  • Peng Tian
    Electrical and Computer Engineering, Northeastern University, Boston, Massachusetts, United States
  • Veysi Yildiz
    Electrical and Computer Engineering, Northeastern University, Boston, Massachusetts, United States
  • Sang Jin Kim
    Ophthalmology, Oregon Health & Science University, Portland, Oregon, United States
    Ophthalmology, Samsung Medical Center, Seoul, Korea (the Republic of)
  • Robison Vernon Paul Chan
    Ophthalmology, University of Illinois at Chicago, Chicago, Illinois, United States
  • Jennifer Dy
    Electrical and Computer Engineering, Northeastern University, Boston, Massachusetts, United States
  • Deniz Erdogmus
    Electrical and Computer Engineering, Northeastern University, Boston, Massachusetts, United States
  • Stratis Ioannidis
    Electrical and Computer Engineering, Northeastern University, Boston, Massachusetts, United States
  • Michael F Chiang
    Ophthalmology and Medical Informatics, Oregon Health & Science University, Portland, Oregon, United States
  • Jayashree Kalpathy-Cramer
    Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, Massachusetts, United States
    Center for Clinical Data Science, Massachusetts General Hospital & Brigham and Women's Hospital, Boston, Massachusetts, United States
Investigative Ophthalmology & Visual Science July 2018, Vol.59, 3938. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      James M Brown, J. Peter Campbell, Susan Ostmo, Peng Tian, Veysi Yildiz, Sang Jin Kim, Robison Vernon Paul Chan, Jennifer Dy, Deniz Erdogmus, Stratis Ioannidis, Michael F Chiang, Jayashree Kalpathy-Cramer; Artificial intelligence in retinopathy of prematurity: development of a fully automated deep convolutional neural network (DeepROP) for plus disease diagnosis. Invest. Ophthalmol. Vis. Sci. 2018;59(9):3938.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : We propose that an image analysis system based on “deep learning”can achieve expert-level performance in an entirely automated fashion.

Methods : A multi-institutional dataset of approximately 6,000 posterior retinal photographs was collected as part of the ongoing “Imaging and Informatics in ROP” (i-ROP) study. Each image was assigned a reference standard diagnosis (RSD) using previously published methods. Images were excluded if two or more image graders labeled them as being “unacceptable for diagnosis”, or if there was evidence of retinal detachment. Retinal photographs were pre-processed using a vessel segmentation algorithm that eliminates variations in image contrast and coloration. For automated diagnosis, the dataset was then split in an 80:20 ratio into five training (n = 4299 ± 70, mean ± standard deviation) and test (n = 1113±70) sets. A deep convolutional neural network (CNN) was trained on each of the training datasets to predict whether an image is ‘normal’, ‘pre-plus’ or ‘plus’. Each of the trained CNNs was evaluated on its corresponding test dataset, with performance measured using receiver operating characteristic (ROC) analysis and areas under the ROC curve (AUC). Image features learned automatically during training were visualized using t-distributed stochastic neighbor embedding (t-SNE) as a two-dimensional scatter plot. The distance between pairs of points on the scatter plot is proportional to their “feature similarity”.

Results : The mean ± standard deviation AUCs for the five models were 0.94±0.01 for diagnosis of ‘pre-plus or worse, and 0.98±0.01 for diagnosis of “plus.” Analysis of image features learned by the CNN using t-SNE shows separation of ‘normal’, ‘pre-plus’ and ‘plus’ images.

Conclusions : The DeepROP CNN can diagnose plus disease from retinal photographs with almost perfect sensitivity and specificity compared to the RSD. t-SNE analysis of image features learned automatically by the CNN suggest a phenotypic continuum of disease severity that could explain inter-observer discrepancies and provide a more granular score for monitoring disease progression.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

Areas under the ROC curve for classification of normal (vs. pre-plus and plus) and plus (vs. normal and pre-plus).

Areas under the ROC curve for classification of normal (vs. pre-plus and plus) and plus (vs. normal and pre-plus).

 

t-SNE plot of retinal images along the disease spectrum, using features extracted from the convolutional neural network.

t-SNE plot of retinal images along the disease spectrum, using features extracted from the convolutional neural network.

×
×

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.

×