Acta Biomedica Scientifica (Dec 2021)

Choroidal neovascularization activity and structure by optical coherence tomography angiography in age related macular degeneration

  • M. A. Kovalevskaya,
  • O. A. Pererva

DOI
https://doi.org/10.29413/ABS.2021-6.6-1.2
Journal volume & issue
Vol. 6, no. 6-1
pp. 12 – 18

Abstract

Read online

Background. In economically developed countries, age-related macular degeneration (AMD) is the leading cause of visual disability among the population of the older age group. The main criterion for the anti-VEGF treatment of neovascular AMD is the activity of choroidal neovascularization (CNV), which is determined by its confi guration. The search for optimal criteria for quantifying the state of the macular region in order to decide on the appointment of anti-VEGF therapy continues.Aim: improving the effi ciency of diagnosis and treatment of AMD based on the assessment of the configuration of vascular system on the “Key to Diagnosis II” platform.Material and methods. The study included 341 patients: 64 % (218 patients, 267 eyes) with non-neovascular AMD, 36 % (123 patients, 174 eyes) – with neovascular AMD. 56 patients (58 eyes) had active type I CNV. Group 1A – active CNV before treatment (9 patients, 9 eyes), group 1B – non-active CNV after treatment with antiVEGF (9 patients, 9 eyes); control group – 10 patients (10 eyes) without AMD. Analysis of OCT-angio images of choriocapillaries included the isolation of CNV, its area, fractal dimension (Df) and the complexity of the vascular system (CVS) counting.Results. Group 1A: Df – 1.5871 ± 0.05, CVS – 2.29 ± 0.29, area – 11734 ± 4866; group 1B: Df – 1.6462 ± 0.08, CVS – 1.65 ± 0.18, area – 6797 ± 3818; control: Df – 1.9167 ± 0.06, CVS – 1, area – 0. Significant differences were found for CVS (p = 0.0003). Df correlates with the CNV area (p = 0.7) and is probably an unreliable parameter due to incomplete visualization of active CNV.Conclusions. CVS is a quantitative biomarker for determining the activity of type 1 CNV in patients with AMD and can serve as a parameter for convolutional neural networks training for automated analysis of OCT angiography images based on the “Key to Diagnosis II” platform

Keywords