June 2020
Volume 61, Issue 7
Free
ARVO Annual Meeting Abstract  |   June 2020
A refined detection scheme and image processing pipeline for multioffset adaptive optics scanning light ophthalmoscopy improves the contrast of retinal ganglion cell layer neurons in humans
Author Affiliations & Notes
  • Elena Gofas Salas
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
  • Min Zhang
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
  • Yuhua Rui
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
  • Valerie Snyder
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
  • Kari V Vienola
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
  • Shanmugathasan Suthaharan
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
    Computer Science, University of North Carolina, Greensboro, North Carolina, United States
  • Ethan A Rossi
    Opthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
    Bioengineering, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
  • Footnotes
    Commercial Relationships   Elena Gofas Salas, None; Min Zhang, None; Yuhua Rui, None; Valerie Snyder, None; Kari Vienola, None; Shanmugathasan Suthaharan, None; Ethan Rossi, University of Rochester (P)
  • Footnotes
    Support  BrightFocus Foundation Grant #G2017082 , NIH CORE Grant P30 EY08098
Investigative Ophthalmology & Visual Science June 2020, Vol.61, 205. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Elena Gofas Salas, Min Zhang, Yuhua Rui, Valerie Snyder, Kari V Vienola, Shanmugathasan Suthaharan, Ethan A Rossi; A refined detection scheme and image processing pipeline for multioffset adaptive optics scanning light ophthalmoscopy improves the contrast of retinal ganglion cell layer neurons in humans. Invest. Ophthalmol. Vis. Sci. 2020;61(7):205.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : We previously showed that retinal ganglion cells (RGCs) can be imaged in vivo in both monkeys and humans with multioffset imaging in adaptive optics scanning light ophthalmoscopy (AOSLO). Though RGCs were visualized with high contrast in monkeys, even revealing sub-cellular structures in some cells, human images had much lower contrast and higher noise. This was likely due to several factors, including: lower light levels for human safety limits, differing detection patterns, fixational eye movements and transverse chromatic aberration between the confocal eye motion reference and multioffset channels. Here we show that a refined multioffset detection scheme and image processing can improve the visualization and quantification of human RGCs in vivo.

Methods : A new optical setup for multioffset detection was implemented to permit confocal and multioffset imaging with the same light source. A pinhole mirror directed the confocal light to a first detector, while the rest of the light passed through and was reimaged to a second retinal conjugate focal plane for multioffset detection. Four healthy subjects (21-40 yrs) were imaged at multiple offsets ranging from 6 to 14 Airy disc diameters (ADD) with a pinhole of ~8 ADD. Image sequences were registered and averaged at each detector position with custom Matlab software. Linear combinations of these averages were computed and the images with the highest contrast of cellular structure were averaged to increase the signal to noise ratio. Finally, an adaptive contrast stretching approach, along with morphological filtering, was deployed to illuminate the perceptual details of RGCs. Cell sizes were measured manually using ImageJ.

Results : We obtained images of the RGC layer mosaic with a soma diameter ranging from 13–20 μm, within the expected range for RGCs. We were able to reproduce images of these cells on the same subject at different time points to prove repeatability of the technique.

Conclusions : A refined detection scheme and image processing for multioffset AOSLO can substantially enhance the contrast of RGC layer somas, permit repeatable imaging of the same cells over time, and improve cell morphometric quantification. Further refinements to this tool could provide a safe, non-invasive, and robust method for quantification of RGC morphology in patients.

This is a 2020 ARVO Annual Meeting abstract.

×
×

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.

×