July 2018
Volume 59, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2018
Automated Detection of Retinal Folds in Papilledema Using En-Face Spectral-Domain Optical Coherence Tomography
Author Affiliations & Notes
  • Jui-Kai Wang
    Electrical and Computer Engineering, University of Iowa, Iowa City, Iowa, United States
  • Jason Agne
    Electrical and Computer Engineering, University of Iowa, Iowa City, Iowa, United States
  • Qingyang Su
    Electrical and Computer Engineering, University of Iowa, Iowa City, Iowa, United States
  • Randy H Kardon
    Center for the Prevention and Treatment of Visual Loss, Iowa City VA Health System, Iowa City, Iowa, United States
    Ophthalmology and Visual Science, The University of Iowa, Iowa City, Iowa, United States
  • Mona K Garvin
    Center for the Prevention and Treatment of Visual Loss, Iowa City VA Health System, Iowa City, Iowa, United States
    Electrical and Computer Engineering, University of Iowa, Iowa City, Iowa, United States
Investigative Ophthalmology & Visual Science July 2018, Vol.59, 2197. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jui-Kai Wang, Jason Agne, Qingyang Su, Randy H Kardon, Mona K Garvin; Automated Detection of Retinal Folds in Papilledema Using En-Face Spectral-Domain Optical Coherence Tomography. Invest. Ophthalmol. Vis. Sci. 2018;59(9):2197.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Optic disc swelling may cause retinal folds that manifest from biomechanical stress and strain. Previously, we published an automated method to detect these folds using spectral-domain optical coherence tomography (SD-OCT) optic-nerve-head (ONH) centered en-face images (Agne et al., OMIA, 2017). In this work, we extended our method to detect retinal folds not only in the area around the ONH but also in the macula. We also improved the proposed approach to identify each fold individually rather than just regionally.

Methods : For each macular and ONH-centered SD-OCT scan pair, four 2D en-face images were created from the entire retinal pigment epithelium (RPE) complex and a thin slab of the internal limiting membrane (ILM), respectively. Then, 107 features were extracted from these en-face images based on Gabor filters, coherence, orientation, and distance from the ONH center (Fig. 1). Twenty SD-OCT ONH and macular pair scans with optic disc swelling from the University of Iowa were used to train a random forest classifier to generate the fold probability maps on both the macular and ONH ILM en-face images. The trained classifier was then applied to the macular and ONH SD-OCT pairs from 27 papilledema eyes in the Idiopathic Intracranial Hypertension Treatment Trial (IIHTT) substudy, and the resulting probability maps from each pair were registered to provide a wider view of the detected folds (Fig. 2, bottom).

Results : The resulting retinal folds in each registered en-face image pair were compared with the folds that were manually outlined. The sensitivity was defined by objects marked as folds in the truth: any pixel found on a connected fold object counted the entire fold as found. The false positive rate was defined in the regular way: any false pixel found to be true. By this criteria, the area under the computed ROC curve was 0.96.

Conclusions : Our method creates the retinal fold probability map in the wide-field ILM en-face image that provides a better chance to visualize the retinal folds between the ONH and macula and further to analyze the stress and strain. Categorizing the types of retinal folds and using fold-related features to distinguish the causes of optic disc swelling constitute future efforts.

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

 

Summary of 107 developed features

Summary of 107 developed features

 

Top: Disc and macula combined ILM en-face image. Middle: Manual fold tracing in red. Bottom: Probability map of detected folds in green.

Top: Disc and macula combined ILM en-face image. Middle: Manual fold tracing in red. Bottom: Probability map of detected folds in green.

×
×

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.

×