June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
Validation of a Novel Approach to Estimate Choroidal Lesion Thickness Using 2D Ultra-widefield Optomap Images
Author Affiliations & Notes
  • Aneesha Ahluwalia
    Byers Eye Institute, Stanford University, Palo Alto, California, United States
  • Michael Yu
    Byers Eye Institute, Stanford University, Palo Alto, California, United States
  • Michael Heiferman
    Byers Eye Institute, Stanford University, Palo Alto, California, United States
  • Edward Korot
    Byers Eye Institute, Stanford University, Palo Alto, California, United States
  • Gina Yu
    Byers Eye Institute, Stanford University, Palo Alto, California, United States
  • Prithvi Mruthyunjaya
    Byers Eye Institute, Stanford University, Palo Alto, California, United States
  • Footnotes
    Commercial Relationships   Aneesha Ahluwalia None; Michael Yu None; Michael Heiferman None; Edward Korot None; Gina Yu None; Prithvi Mruthyunjaya None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 908. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Aneesha Ahluwalia, Michael Yu, Michael Heiferman, Edward Korot, Gina Yu, Prithvi Mruthyunjaya; Validation of a Novel Approach to Estimate Choroidal Lesion Thickness Using 2D Ultra-widefield Optomap Images. Invest. Ophthalmol. Vis. Sci. 2023;64(8):908.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Thickness of melanocytic choroidal lesions (MCL) ≥2 mm is an important risk factor for malignant potential. Our group previously described the novel use of green channel 2D ultra-widefield (UWF) Optomap images in lieu of B-scan ultrasonography to identify lesions with thickness ≥2 mm with high sensitivity and specificity. Here, we performed an internal validation of our previous model and predicted the thickness of MCL using pixel intensity.

Methods : In our prior work, we analyzed 153 MCL from 153 patients seen by the Ocular Oncology Service at the Byers Eye Institute (Stanford University, Palo Alto, CA). This study included a cohort of 25 MCL from 25 patients that were not included in previous analyses. All patients underwent 2D UWF fundus imaging and B-scan ultrasonography. Pixel intensity difference (PID) was calculated from green channel UWF photographs by subtracting the pixel intensity of the surrounding retina from the pixel intensity within the lesion, as previously described. Using our prior linear regression model, PID was then correlated to lesion thickness to identify lesions ≥2 mm using a threshold of PID ≥5.

Results : Of 25 included MCL, 7 had thickness ≥2 mm. The mean thickness of included lesions was 1.94 ± 2.67 mm (median: 1.06; range: 0.48 to 12.74). Mean calculated PID was 6.47 ± 14.55 mm (median: 1.9; range: -6.9 to 50.1). There were 9 lesions above the PID threshold ≥5, indicating an estimated lesion thickness ≥2 mm. This corresponded with a sensitivity of 100% and specificity of 89% in identifying lesions with thickness ≥2 mm. Next, PID was used to estimate thickness of MCL, utilizing a linear model derived from our prior study, which was compared to lesion thickness as measured by B-scan ultrasonography. For lesions ≥2 mm, the mean percentage difference between PID-predicted thickness and B-scan thickness was 23%. For lesions <2 mm, the mean absolute difference between PID-predicted and B-scan thickness was 0.29 ± 0.19 mm.

Conclusions : This study validates that the relative PID of MCL on 2D UWF fundus imaging could be used to reliably identify lesions with thickness ≥2 mm. PID could secondarily function as a proxy for estimating lesion thickness in the absence of B-scan ultrasonography. With further validation, automated measurements of lesion relative pixel intensity may serve as a more readily available metric in evaluating lesion malignant potential.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

×
×

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.

×