June 2017
Volume 58, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2017
Deep learning for the identification of plus disease in retinopathy of prematurity
Author Affiliations & Notes
  • Jayashree Kalpathy-Cramer
    MGH/Harvard Medical School, Charlestown, Massachusetts, United States
  • J. Peter Campbell
    Ophthalmology, Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, United States
  • Sang Kim
    Ophthalmology, Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, United States
    Ophthalmology, Samsung Medical Center, Seoul, Korea (the Republic of)
  • Ryan Swan
    Ophthalmology, Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, United States
  • Karyn Elizabeth Jonas
    Ophthalmology, University of Illinois, Chicago, Chicago, Illinois, United States
  • Susan Ostmo
    Ophthalmology, Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, United States
  • Peng Tian
    Northeastern University, Boston, Massachusetts, United States
  • Dharanish Kedarisetti
    Northeastern University, Boston, Massachusetts, United States
  • Stratis Ioannidis
    Northeastern University, Boston, Massachusetts, United States
  • Deniz Erdogmus
    Northeastern University, Boston, Massachusetts, United States
  • RV Paul Chan
    Ophthalmology, University of Illinois, Chicago, Chicago, Illinois, United States
  • Michael F Chiang
    Ophthalmology, Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, United States
  • Footnotes
    Commercial Relationships   Jayashree Kalpathy-Cramer, None; J. Peter Campbell, None; Sang Kim, None; Ryan Swan, None; Karyn Jonas, None; Susan Ostmo, None; Peng Tian, None; Dharanish Kedarisetti, None; Stratis Ioannidis, None; Deniz Erdogmus, None; RV Paul Chan, Visunex (C); Michael Chiang, Clarity Medical Systems (S), Novartis (C)
  • Footnotes
    Support  Supported by grant P30EY10572 from the National Institutes of Health (Bethesda, MD), and by unrestricted departmental funding from Research to Prevent Blindness (New York, NY). R01EY19474 and R21EY22387 from the National Institutes of Health (Bethesda, MD), and by grant 1622679 from the National Science Foundation (Arlington, VA).
Investigative Ophthalmology & Visual Science June 2017, Vol.58, 5554. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Jayashree Kalpathy-Cramer, J. Peter Campbell, Sang Kim, Ryan Swan, Karyn Elizabeth Jonas, Susan Ostmo, Peng Tian, Dharanish Kedarisetti, Stratis Ioannidis, Deniz Erdogmus, RV Paul Chan, Michael F Chiang; Deep learning for the identification of plus disease in retinopathy of prematurity. Invest. Ophthalmol. Vis. Sci. 2017;58(8):5554.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Retinopathy of Prematurity (ROP) is leading cause of childhood blindness worldwide. In the United States and globally, there is a shortage of adequately trained clinicians who can manage ROP at point of care. Identification of plus disease is key component of the diagnostic process. However, plus disease classification has been shown to be subjective and qualitative. This study explores the use of deep learning to identify infants with clinically significant ROP and to calculate a severity score.

Methods : We compared three deep learning approaches to ROP disease classification. We used two datasets for the experiments. Our smaller, highly characterized dataset of 195 images had manual segmentations of the vasculature, plus disease classification, and disease severity rankings based on pairwise comparisons. Our larger dataset consisted of over 6000 images (multiple views) from 650 patients, including 2400 images of the posterior pole. The larger set has a 3 level classification provided by 3-5 experts.

We utilized a deep learning approach with a U-network architecture to develop a segmentation algorithm. We then extracted quantitative measure of the vascularity from the segmented images and created a random forest classifier for 3 level disease classification (plus, pre-plus, normal). Neighborhood approximate forests were used to create a severity score. Finally, neural networks were developed to perform image classification without segmentation. Data pre-processing and augmentation were employed. We calculated the area under the receiver operating characteristic curve (AUC) for the deep learning algorithm, as well as correlation coefficients (CC) for the algorithm rankings versus pairwise comparison rankings.

Results : A deep learning system for vascular segmentation based on a U-network architecture had an AUC of 0.97. Features extracted from the segmented images are able to classify images into 3 classes and correlated well with the severity ranking (CC 0.78, p<0.05).

Conclusions : A deep learning system for vascular segmentation based on a U-network architecture is extremely effective for classifying plus disease in ROP diagnosis. Deep learning has great potential in automated analysis of ROP.

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

×
×

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.

×