June 2020
Volume 61, Issue 7
Free
ARVO Annual Meeting Abstract  |   June 2020
Use of automated processing of en face swept-source optical coherence tomography images to evaluate vascular disease severity in retinopathy of prematurity.
Author Affiliations & Notes
  • Kai Seely
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Vincent Tai
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • S. Grace Prakalapakorn
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Sara Grace
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Stephanie J Chiu
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Sharon Freedman
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Cynthia Toth
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Footnotes
    Commercial Relationships   Kai Seely, Duke University (P); Vincent Tai, Duke University (P); S. Prakalapakorn, None; Sara Grace, None; Stephanie Chiu, Duke University (P); Sharon Freedman, None; Cynthia Toth, Alcon (F), Duke University (P), EMMES (C), Hemosonics (F)
  • Footnotes
    Support  NIH/NEI EY025009, NIH/NEI EY028227, NIH/NEI EY005722, NIH/NCATS TR002555, Research supported by the Research to Prevent Blindness Stein Innovation Award
Investigative Ophthalmology & Visual Science June 2020, Vol.61, 2156. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Kai Seely, Vincent Tai, S. Grace Prakalapakorn, Sara Grace, Stephanie J Chiu, Sharon Freedman, Cynthia Toth; Use of automated processing of en face swept-source optical coherence tomography images to evaluate vascular disease severity in retinopathy of prematurity.. Invest. Ophthalmol. Vis. Sci. 2020;61(7):2156.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Preterm infants at risk for retinopathy of prematurity (ROP) require screening for the development of severe disease. Swept-source optical coherence tomography (SS-OCT) may be an alternative to current screening methods. We describe a novel method for creating high-quality en face SS-OCT images and assess the feasibility and efficacy of using these images for evaluating vascular disease in preterm infants at risk for ROP.

Methods : In this prospective study, we imaged preterm infants using investigational, handheld, bedside SS-OCT. We selected a stratified random sample of 17 eyes with a range of vascular disease severity on clinical exam (6 no plus, 4 pre-plus, 7 plus). For each eye, we selected the highest-quality imaging session; auto-segmented the images using custom software (DOCTRAP v63.9), optimizing for visualization of retinal vessels and elimination of choroidal vasculature; and auto-montaged (i2k Retina, DualAlign) to increase field of view. Three masked ROP experts reviewed the resulting en face images (Fig 1) for quality assessment, grading of no, pre-, or plus disease, and ranking of vascular disease severity. The modal (consensus) vascular disease severity ranking was used as a summary score.

Results : Processed images were gradable for no, pre-, or plus disease in 15 (88%) and rankable for vascular severity in 17 (100%) of 17 eyes. Intergrader agreement for grading was moderate (weighted kappa 0.67; 95% CI 0.42-0.86). Intergrader agreement for ranking was excellent (ICC 0.98; 95% CI 0.96-0.99) (Fig 2). Sensitivity and specificity of the grading were 61.1% and 92.6%, respectively, for detecting plus disease and 96.3% and 100%, respectively, for detecting pre-plus or plus disease.

Conclusions : These pilot data, based on auto-segmented and auto-montaged en face SS-OCT images with enhanced vessel contrast and field of view, support the feasibility of using processed SS-OCT images to detect ROP vascular disease in preterm infants. Larger follow-up studies are needed to evaluate their utility as a potential alternative to current ROP screening methods.

This is a 2020 ARVO Annual Meeting abstract.

 

Fig 1. Unprocessed (left) and auto-segmented and auto-montaged (right) SS-OCT images of ROP eyes with no plus (A), pre-plus (B), and plus disease (C).

Fig 1. Unprocessed (left) and auto-segmented and auto-montaged (right) SS-OCT images of ROP eyes with no plus (A), pre-plus (B), and plus disease (C).

 

Fig 2. Vascular disease severity rankings for 17 eyes exhibit excellent intergrader agreement and correlate with clinical ROP grade.

Fig 2. Vascular disease severity rankings for 17 eyes exhibit excellent intergrader agreement and correlate with clinical ROP grade.

×
×

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.

×