Quantifying disrupted outer retinal-subretinal layer in SD-OCT images in choroidal neovascularization

Area of Science:

• Ophthalmology and Medical Imaging
• Biomedical Engineering
• Computational Imaging

Background:

• Exudative age-related macular degeneration (eAMD) involves choroidal neovascularization (CNV), leading to fluid accumulation.
• Accurate measurement of the outer retinal-subretinal (ORSR) layer is crucial for managing eAMD.
• Current methods for ORSR layer segmentation in spectral-domain optical coherence tomography (SD-OCT) are often inaccurate or manual.

Purpose of the Study:

• To develop and validate a fully automated method for quantifying ORSR layer thickness in SD-OCT scans.
• To assess the accuracy and reproducibility of the automated method in patients with CNV due to eAMD.

Main Methods:

• A 3-dimensional (3D) graph-search approach with an adaptive cost function was employed for automated segmentation.
• Volumetric SD-OCT scans from 23 subjects with CNV were analyzed.
• Segmentation accuracy was graded by a retinal specialist, and reproducibility was assessed using root mean square coefficient of variation (RMS CV).

Main Results:

• The automated adaptive method achieved 21/23 generally accurate segmentations, compared to only 1/23 with a standard approach.
• The average ORSR thickness in CNV patients (78.0 μm) was significantly greater than in normal subjects (51.5 μm).
• The automated method demonstrated good reproducibility with an RMS CV of 8.1%.

Conclusions:

• A fully automated 3D method for segmenting the ORSR layer in SD-OCT of eAMD patients with CNV has been successfully developed.
• This method accurately quantifies ORSR layer thickness even in the presence of fluid.
• The technique has the potential to improve the accuracy and efficiency of anti-VEGF treatment management for eAMD.