June 2022
Volume 63, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2022
Comparison between Drusen Volume Calculations on Heidelberg OCT using Automated vs Manually Adjusted Segmentation for Intermediate AMD and Its Association with Retinal Pigment Epithelium Bruch’s Membrane Thickness
Author Affiliations & Notes
  • Thomas Koch
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
    Harvard Retina Imaging Lab, Boston, Massachusetts, United States
  • Archana Nigalye
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
    Harvard Retina Imaging Lab, Boston, Massachusetts, United States
  • Shrinivas Pundlik
    Department of Ophthalmology Harvard Medical School, Schepens Eye Research Institute of Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Gregory Tsougranis
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Raviv Katz
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
    Harvard Retina Imaging Lab, Boston, Massachusetts, United States
  • Itika Garg
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
    Harvard Retina Imaging Lab, Boston, Massachusetts, United States
  • Hannah Wescott
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
    Harvard Retina Imaging Lab, Boston, Massachusetts, United States
  • Deeba Husain
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • John B Miller
    Retina Service, Massachusetts Eye and Ear, Boston, Massachusetts, United States
    Harvard Retina Imaging Lab, Boston, Massachusetts, United States
  • Footnotes
    Commercial Relationships   Thomas Koch None; Archana Nigalye None; Shrinivas Pundlik None; Gregory Tsougranis None; Raviv Katz None; Itika Garg None; Hannah Wescott None; Deeba Husain Allergan, Code C (Consultant/Contractor), Genentech, Code C (Consultant/Contractor), Omeicos Therapeutics, Code C (Consultant/Contractor), National Eye Institute, Code F (Financial Support), Lions VisionGift, Code F (Financial Support), Commonwealth Grant, Code F (Financial Support), Lions International, Code F (Financial Support), Syneos LLC, Macular Society, Code F (Financial Support); John Miller Alcon, Code C (Consultant/Contractor), Allergan, Code C (Consultant/Contractor), Carl Zeiss, Code C (Consultant/Contractor), Sunovion, Code C (Consultant/Contractor), Genentech, Code C (Consultant/Contractor)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2022, Vol.63, 1052 – F0299. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Thomas Koch, Archana Nigalye, Shrinivas Pundlik, Gregory Tsougranis, Raviv Katz, Itika Garg, Hannah Wescott, Deeba Husain, John B Miller; Comparison between Drusen Volume Calculations on Heidelberg OCT using Automated vs Manually Adjusted Segmentation for Intermediate AMD and Its Association with Retinal Pigment Epithelium Bruch’s Membrane Thickness. Invest. Ophthalmol. Vis. Sci. 2022;63(7):1052 – F0299.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Drusen burden is a key indicator of age-related macular degeneration (AMD) severity. Real time manual measurement of drusen volume (DV) by clinicians is not practical and automated estimates with inbuilt software save time, but are not readily available. Existing work on DV has shown variation between OCT devices and with automated vs. manually adjusted segmentation. We further examined this variation.

Methods : Retrospective, cross-sectional study of intermediate AMD eyes imaged with macula centered 97 line enhanced depth imaging OCT (Spectralis, Heidelberg). Volume scans were processed for automated segmentation of the RPE and BM to generate the drusen volume. To manually adjust segmentation, RPE and BM were fit to demarcation in each B-scan prior to calculating DV. The average retinal thickness (RT) and DV, and highest retinal pigment epithelium/Bruch’s Membrane thickness (RPEBMt) were measured in all 9 ETDRS grid sectors. Multi-level linear regression models, accounting for participant and eye-in-participant levels, were used to associate the log RT and log DV measurements with the processing method (automated vs. manual), RPEBMt(continuous measure), and their interaction.

Results : We included 10 eyes of 5 participants. Over entire ETDRS grid, the mean difference between manual and automated processing for RT and DV was 7.9±21.1 μm and 0.015±0.04 mm3, respectively. Larger differences between manual and automated methods were observed for inner grid segments (GS) and the center than the outer GS for both RT (Inner: 12.3±26.6 μm; Outer: 2.5±8.4 μm) and DV (Inner: 0.017±0.037 mm3; Outer: 0.012±0.044 mm3). The mean±SD of RPEBMt parameter was 85.3±51.9μ. Significant interaction seen between RPEBMt and processing method for both RT (β=0.006, p<0.001) and DV (β=0.007, p<0.001), meaning the variation between manual and automated processing became larger as the RPEBMt increased.

Conclusions : Presence of large drusen affected the automated segmentation resulting in larger deviations, as measured by significant association between RPEBMt and DV and RT, compared to the manual segmentation. In eyes with large drusen, the current automated segmentation does not reliably measure these parameters, and there is a need for manual review to mitigate the effects of segmentation compromise.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

×
×

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.

×