Automatic choroidal segmentation in OCT images using supervised deep learning methods

Abstract

The analysis of the choroid in the eye is crucial for our understanding of a range of ocular diseases and physiological processes. Optical coherence tomography (OCT) imaging provides the ability to capture highly detailed cross-sectional images of the choroid yet only a very limited number of commercial OCT instruments provide methods for automatic segmentation of choroidal tissue. Manual annotation of the choroidal boundaries is often performed but this is impractical due to the lengthy time taken to analyse large volumes of images. Therefore, there is a pressing need for reliable and accurate methods to automatically segment choroidal tissue boundaries in OCT images. In this work, a variety of patch-based and fully-convolutional deep learning methods are proposed to accurately determine the location of the choroidal boundaries of interest. The effect of network architecture, patch-size and contrast enhancement methods was tested to better understand the optimal architecture and approach to maximize performance. The results are compared with manual boundary segmentation used as a ground-truth, as well as with a standard image analysis technique. Results of total retinal layer segmentation are also presented for comparison purposes. The findings presented here demonstrate the benefit of deep learning methods for segmentation of the chorio-retinal boundary analysis in OCT images.

Retinal Boundary Segmentation in Stargardt Disease Optical Coherence Tomography Images Using Automated Deep Learning

“…We have developed this fully semantic network and graph search (FS-GS) in our previous work and have previously demonstrated its application to retinal and choroidal segmentation in images with no pathologic changes. 22 Additionally, we have highlighted the ability of training such a network to be noise resilient when provided with OCT images of poor quality. 43 As was the case in our previous studies, eight filters were used in the initial set of convolution blocks with this doubled after each subsequent pooling layer.…”

“…This approach is similar to previous studies that have reported related segmentation performance improvements using a similar technique. 20 , 22 An example of the application of the Girard filter is shown in Figure 1 . To boost the diversity within the dataset, data augmentation methods were also used.…”

“…The GS method used for boundary delineation is similar to that used in a number of previous studies. 12 , 18 – 20 , 22 Using a trained network, layer probability maps for test images may be obtained. To obtain boundary probability maps, edge detection is then performed with a boundary probability map formed for each boundary.…”

“…The first is a patch-based machine learning method. 18 – 20 , 21 , 22 For this, we use the Cifar CNN architecture and a set of patch classes (single background) proposed by Fang et al 18 and a graph search method and 64 × 32 patch size used by Kugelman et al 20 Here, patches are constructed using the same set of images as the proposed method, with one patch for each class sampled from every column where segmentations are present. The other two methods for comparison, not based on machine learning, are publicly available tools including the OCTExplorer tool (part of the IOWA Reference Algorithms [Retinal Image Analysis Lab, Iowa Institute for Biomedical Imaging, Iowa City, IA]), 14 , 48 , 49 and the AURA tool for retinal layer segmentation.…”

“…Early versions of these methods were built around standard image processing techniques and algorithms. 9 – 14 More recently, machine learning and deep learning methods have been used, including support vector machines, 15 , 16 random forest classifiers, 17 patch-based classification with convolutional neural networks 18 – 22 or recurrent neural networks, 20 , 22 semantic segmentation with fully convolutional (encoder–decoder) networks, 22 – 26 and other deep learning methods. 27 – 30 Importantly, some of these methods have been applied to OCT images from patients with age-related macular degeneration, 18 , 20 , 24 , 27 diabetic retinopathy, 11 , 25 macular telangiectasia type 2, 29 diabetic macular oedema, 13 , 23 , 24 pigment epithelium detachment, 28 glaucoma, 15 , 30 multiple sclerosis 17 , 26 retinitis pigmentosa, 31 and neurodegenerative diseases.…”