ReLayer: a free, online tool for measuring retinal thickness based on a new automatic method for the segmentation of layers in OCT images.

Abstract

Purpose : To provide a free, online tool for automatically segmenting OCT images and computing summary measures such as retinal thickness.

Methods : ReLayer (https://relayer.online) is an online platform to which OCT scan images can be uploaded and analyzed (Figure 1). Results can be downloaded in text files. The segmentation method is based on the detection of horizontal edges, and includes a novel, 1-dimensional active contour model, designed to locate the inner limiting membrane (ILM) and retinal pigment epithelium (RPE). The method was optimized for B-scans from Topcon (Tokyo, Japan) 3D OCT-2000 and Heidelberg Engineering (Heidelberg, Germany) Spectralis. The method was tested against the manufacturer’s segmentation (Reference Standard; RS) on 48 B-scans acquired with Spectralis from 48 healthy eyes, and the results were validated by a clinical expert.

Results : Segmentation of a single B-scan takes ≤ 1 second. Average difference in retinal thickness (standard deviation) from ReLayer and RS was 2.46 (0.83) micrometers, an average value below the resolution of the instrument (3.87 micrometers). The clinical validation assessed that both segmentations were correct in 100% of test images (Figure 2).

Conclusions : ReLayer is designed to be a free platform for automatic segmentation of OCT scans. Segmentation of ILM and RPE in ReLayer is equivalent to that acquired by software on Spectralis OCT in healthy eyes.ReLayer could represent a valuable tool for researchers requiring the full segmentation across all scans: data which are often not made available by commercial software. Future developments will include the segmentation of new layers, of scans from other manufacturers and of tools for manual adjustments and batch processing.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

A screenshot of the web-based interface of ReLayer accessible at https://relayer.online.

 

The segmentation results from Reference Standard (green) and from ReLayer (dashed red) in a subset of the 48 test images.