Monocular Retinal Depth Estimation and Joint Optic Disc and Cup Segmentation using Adversarial Networks

TL;DR

This paper introduces a novel adversarial network approach for estimating retinal depth from a single RGB image, enabling joint optic disc and cup segmentation to aid glaucoma assessment, with high accuracy demonstrated on a public dataset.

Contribution

The paper presents a new adversarial network method for monocular retinal depth estimation and its application to joint optic disc and cup segmentation, outperforming existing techniques.

Findings

Achieved a correlation coefficient of 0.92 in depth estimation

Outperformed state-of-the-art methods in accuracy

Enabled effective joint segmentation of optic disc and cup

Abstract

One of the important parameters for the assessment of glaucoma is optic nerve head (ONH) evaluation, which usually involves depth estimation and subsequent optic disc and cup boundary extraction. Depth is usually obtained explicitly from imaging modalities like optical coherence tomography (OCT) and is very challenging to estimate depth from a single RGB image. To this end, we propose a novel method using adversarial network to predict depth map from a single image. The proposed depth estimation technique is trained and evaluated using individual retinal images from INSPIRE-stereo dataset. We obtain a very high average correlation coefficient of 0.92 upon five fold cross validation outperforming the state of the art. We then use the depth estimation process as a proxy task for joint optic disc and cup segmentation.