Investigative Ophthalmology & Visual Science Cover Image for Volume 65, Issue 7
June 2024
Volume 65, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2024
AI-based correlation of functional retinal testing with retinal fluid volumes and compartments in neovascular age-related macular degeneration
Author Affiliations & Notes
  • Laetitia Hinterhuber
    Medizinische Universitat Wien, Wien, Wien, Austria
  • Gregor Sebastian Reiter
    Department of Ophthalmology and Optometry, Medizinische Universitat Wien, Wien, Wien, Austria
  • Klaudia Kostolna
    Department of Ophthalmology and Optometry, Medizinische Universitat Wien, Wien, Wien, Austria
  • Hamza Mohamed
    Medizinische Universitat Wien, Wien, Wien, Austria
  • Leonard Mana Coulibaly
    Department of Ophthalmology and Optometry, Medizinische Universitat Wien, Wien, Wien, Austria
  • Philipp Fuchs
    Department of Ophthalmology and Optometry, Medizinische Universitat Wien, Wien, Wien, Austria
  • Simon Schürer-Waldheim
    Laboratory for Ophthalmic Image Analysis, Medizinische Universitat Wien, Wien, Wien, Austria
  • Markus Gumpinger
    Laboratory for Ophthalmic Image Analysis, Medizinische Universitat Wien, Wien, Wien, Austria
  • Hrvoje Bogunovic
    Laboratory for Ophthalmic Image Analysis, Medizinische Universitat Wien, Wien, Wien, Austria
  • Oliver Leingang
    RetInSight, Vienna, Austria, Austria
  • Ursula Schmidt-Erfurth
    Department of Ophthalmology and Optometry, Medizinische Universitat Wien, Wien, Wien, Austria
  • Footnotes
    Commercial Relationships   Laetitia Hinterhuber None; Gregor Reiter RetInSight, Code F (Financial Support); Klaudia Kostolna None; Hamza Mohamed None; Leonard Coulibaly None; Philipp Fuchs None; Simon Schürer-Waldheim None; Markus Gumpinger None; Hrvoje Bogunovic Heidelberg Engineering, Code F (Financial Support); Oliver Leingang RetInSight, Code E (Employment); Ursula Schmidt-Erfurth Heidelberg Engineering, Code C (Consultant/Contractor), Heidelberg Engineering, Code F (Financial Support), RetInSight, Code F (Financial Support)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2024, Vol.65, 3259. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Laetitia Hinterhuber, Gregor Sebastian Reiter, Klaudia Kostolna, Hamza Mohamed, Leonard Mana Coulibaly, Philipp Fuchs, Simon Schürer-Waldheim, Markus Gumpinger, Hrvoje Bogunovic, Oliver Leingang, Ursula Schmidt-Erfurth; AI-based correlation of functional retinal testing with retinal fluid volumes and compartments in neovascular age-related macular degeneration. Invest. Ophthalmol. Vis. Sci. 2024;65(7):3259.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Microperimetry (MP) is emerging as a clinical endpoint in retinal disease. We aimed to investigate the spatial relation between retinal function and morphology on optical coherence tomography (OCT) imaging supported by deep-learning (DL)-algorithms, whilst assessing the inter- and intradevice repeatability of MP in neovascular age-related macular degeneration (nAMD).

Methods : 40 eyes of 40 patients, 20 with nAMD and 20 healthy controls completed two consecutive MP examinations on MP3 and MAIA after eligibility was confirmed with OCT (Spectralis, 97 B-scans). A DL-algorithm, The Fluid Monitor (RetInSight) quantified intraretinal, subretinal fluid (IRF, SRF) and pigment epithelial detachment (PED) after all OCT B-scans were superimposed with 45 stimuli from the MP grid. The impact of fluid on sensitivity was analyzed via multivariate mixed effect models. Repeatability was assessed with intraclass correlation coefficient (ICC), Bland-Altman plots and coefficient of repeatability (CoR) over stimuli with fluid.

Results : 7.200 MP stimuli were analyzed. Mean sensitivity in nAMD and controls was 23±6 and 29±2dB in MP3 and 20±6 and 25±3dB in MAIA, respectively. Overall sensitivity showed significant differences in nAMD compared to controls (p<0.001). IRF and PED had a significant negative effect on sensitivity with estimates of -3dB/nl (p<0.001) in MP3 and -1dB/nl in MAIA (p<0.004) and -2dB/nl (p<0.001) on both devices, respectively. SRF had no significant impact on function in MP3 (p=0.09), but reached significance in MAIA (-1dB/nl,p=0.002). Repeatability in MAIA was worse in stimuli over IRF (CoR ±8dB) compared to SRF, PED (both ±7dB) and comparable between all fluid compartments in MP3 with CoR ±7dB for both IRF, SRF and ±6dB for PED. ICC was excellent for intra, -and interdevice measurements with 0.93 and 0.94 for consecutive runs and 0.87 and 0.92 between devices.

Conclusions : DL allows for exact quantification of fluid volume together with point-wise topographic correlation to retinal function. Point-wise correlation showed significantly impaired sensitivity in nAMD compared to controls, with IRF and PED as important predictors of reduced function. Assessment of repeatability in MP enhances the trustworthiness of MP as a functional endpoint in clinical trials and provides clinical evidence for resolution of retinal fluid by compartment and volume.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

 

×
×

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.

×