July 2019
Volume 60, Issue 9
Free
ARVO Annual Meeting Abstract  |   July 2019
Quantification of choriocapillaris structure in high-resolution OCTA images
Author Affiliations & Notes
  • Brennan Marsh-Armstrong
    Ophthalmology, University of California Davis, Davis, California, United States
  • Justin Migacz
    Ophthalmology, University of California Davis, Davis, California, United States
  • Ravi Jonnal
    Ophthalmology, University of California Davis, Davis, California, United States
  • John S Werner
    Ophthalmology, University of California Davis, Davis, California, United States
  • Footnotes
    Commercial Relationships   Brennan Marsh-Armstrong, None; Justin Migacz, None; Ravi Jonnal, None; John Werner, None
  • Footnotes
    Support  R01 EY024239, P30 EY012576, T32 EY15387, R00 EY026068
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 3089. doi:https://doi.org/
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Brennan Marsh-Armstrong, Justin Migacz, Ravi Jonnal, John S Werner; Quantification of choriocapillaris structure in high-resolution OCTA images. Invest. Ophthalmol. Vis. Sci. 2019;60(9):3089. doi: https://doi.org/.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Analysis of choriocapillaris vascular morphology may provide a novel avenue for studying and diagnosing a range of retinal pathologies. Megahertz-rate OCT angiography has recently produced the first OCT choriocapillaris images with consistently resolved vessels without adaptive optics. The purpose of this study is to develop metrics for quantifying numerous parameters of choriocapillaris anatomy through the analysis of these high-contrast images.

Methods : Healthy subjects were imaged 6 degrees nasal from the fovea with a custom OCT system using a Fourier-domain mode-locked laser operating at a 1.7 MHz A-scan rate. The flattened en face images of the choriocapillaris layer (~10 µm thick) were extracted and averaged. These images underwent noise-reduction, skeletonization, and were then mapped as a graph with vessel branching-points as nodes. To overcome the inconsistent local intensity of the image a local-min/max normalization algorithm was developed. With the masked images we quantified choriocapillaris vessel diameter and flow void distance-to-nearest-capillary. From the vessel trace, we quantified the distance between vessel branch points and branching factor.

Results : For the subject being displayed, a flattened choriocapillaris image with capillary-level resolution was processed(Fig 1A), traced (Fig 1B), and inverted to produce a flow void map (Fig 2A). Both the trace and flow void map were quantified. The average branch-to-branch distance of vessels was 16(7) µm. Each vasculature branching point had an average of 2.4(7) branching vessels. Locations within flow voids are an average of 8.4(4.7) µm to the nearest vessel center (Fig 2B). Vessels were observed to be 7.2(2.9) µm in radius on average.

Conclusions : Tracing-based quantification of high-resolution OCT images of the choriocapillaris is a new tool for precise analysis of the vasculature underlying the retina. This approach, for which we have developed an imaging system and computational analysis technique, offers a novel avenue for assessing retinal vasculature and vasculature-affecting diseases including glaucoma and age-related macular degeneration.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Figure 1: A) A normal subject’s en face choriocapillaris OCTA image was B) skeletonized to identify vasculature patterning.

Figure 1: A) A normal subject’s en face choriocapillaris OCTA image was B) skeletonized to identify vasculature patterning.

 

Figure 2: The vessel trace was used to create a A) heatmap of flow voids colored by distance-to-nearest-vessel, and B) a histogram of flow void radius across the image.

Figure 2: The vessel trace was used to create a A) heatmap of flow voids colored by distance-to-nearest-vessel, and B) a histogram of flow void radius across the image.

×
×

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.

×