Segmentation of Retinal Low-Cost Optical Coherence Tomography Images using Deep Learning

TL;DR

Retinal scans of a novel self-examination low-cost full-field OCT (SELF-OCT) are segmented using a deep learningbased approach, and it is shown that the CNN-based approach can segment the retina with high accuracy, whereas the segmentation of the PED proves to be challenging.

Abstract

The treatment of age-related macular degeneration (AMD) requires continuous eye exams using optical coherence tomography (OCT). The need for treatment is determined by the presence or change of disease-specific OCT-based biomarkers. Therefore, the monitoring frequency has a significant influence on the success of AMD therapy. However, the monitoring frequency of current treatment schemes is not individually adapted to the patient and therefore often insufficient. While a higher monitoring frequency would have a positive effect on the success of treatment, in practice it can only be achieved with a home monitoring solution. One of the key requirements of a home monitoring OCT system is a computer-aided diagnosis to automatically detect and quantify pathological changes using specific OCT-based biomarkers. In this paper, for the first time, retinal scans of a novel self-examination low-cost full-field OCT (SELF-OCT) are segmented using a deep learning-based approach. A convolutional neural network (CNN) is utilized to segment the total retina as well as pigment epithelial detachments (PED). It is shown that the CNN-based approach can segment the retina with high accuracy, whereas the segmentation of the PED proves to be challenging. In addition, a convolutional denoising autoencoder (CDAE) refines the CNN prediction, which has previously learned retinal shape information. It is shown that the CDAE refinement can correct segmentation errors caused by artifacts in the OCT image.

Figures (4)

• Figure 1: Comparison of two retinal scans of the same AMD patient with PED between Spectralis OCT from Heidelberg Engineering (left) and SELF-OCT (right).
• Figure 2: Schema of the segmentation pipeline. First row: multi-class segmentation of the OCT volume with the U-Net and generation of the retinal shape via binarization. Second row: Shape refinement by the CDAE and fusion with the U-Net segmentation.
• Figure 3: Architecture of the proposed 3D U-Net. The notation C$_1$/C$_2$@W$\times$H$\times$D at each blocks describes the number of output channels of the first or second convolution (C$_{1/2}$) for a given image resolution (width$\times$height$\times$depth).
• Figure 4: Segmentation results of the proposed approach of two retinal B-scans of two individual patients with retina () and PED (): B-scan (a)/(e), ground truth (b)/(f), U-Net segmentation (c)/(g), CDAE segmentation (d)/(h). Areas with low SNR or motion artifacts (white arrow) (a)/(e) can lead to a false segmentation (c)/(g) and are corrected by the CDAE refinement (d)/(h).