Epiretinal Membrane Detection at the Ophthalmologist Level using Deep Learning of Optical Coherence Tomography

Abstract

Purpose: Previous deep learning studies on optical coherence tomography (OCT) mainly focused on diabetic retinopathy and age-related macular degeneration. We proposed a deep learning model that can identify epiretinal membrane (ERM) in OCT with ophthalmologist-level performance. Design: Cross-sectional study. Participants: A total of 3,618 central fovea cross section OCT images from 1,475 eyes of 964 patients. Methods: We retrospectively collected 7,652 OCT images from 1,197 patients. From these images, 2,171 were normal and 1,447 were ERM OCT. A total of 3,141 OCT images was used as training dataset and 477 images as testing dataset. DL algorithm was used to train the interpretation model. Diagnostic results by four board-certified non-retinal specialized ophthalmologists on the testing dataset were compared with those generated by the DL model. Main Outcome Measures: We calculated for the derived DL model the following characteristics: sensitivity, specificity, F1 score and area under curve (AUC) of the receiver operating characteristic (ROC) curve. These were calculated according to the gold standard results which were parallel diagnoses of the retinal specialist. Performance of the DL model was finally compared with that of non-retinal specialized ophthalmologists. Results: Regarding the diagnosis of ERM in OCT images, the trained DL model had the following characteristics in performance: sensitivity: 98.7%, specificity: 98.0%, and F1 score: 0.945. The accuracy on the training dataset was 99.7% (95% CI: 99.4 - 99.9%), and for the testing dataset, diagnostic accuracy was 98.1% (95% CI: 96.5 - 99.1%). AUC of the ROC curve was 0.999. The DL model slightly outperformed the average non-retinal specialized ophthalmologists. Conclusions: An ophthalmologist-level DL model was built here to accurately identify ERM in OCT images. The performance of the model was slightly better than the average non-retinal specialized ophthalmologists. The derived model may play a role to assist clinicians to promote the efficiency and safety of healthcare in the future.

Epiretinal Membrane Detection in Optical Coherence Tomography Retinal Images Using Deep Learning

“…Baamonde et al [16], [18] and Fang et al [47] reported methods based on conventional feature extraction and machine learning. Lo et al [20] and Lu et al [17] proposed deep learning methods to identify the presence of ERM. Finally, Sonobe et al [19] compared support vector machines (SVM) and deep learning techniques using the reconstructed surface of the retina from OCT images to detect ERM.…”

“…Similarly, Lo et al [20] proposed a method to detect ERM at a medical specialist level. A ResNet-101 architecture was employed with a dataset comprising 3,618 OCT images (2,171 normal and 1,447 ERM).…”

“…The development of fully automatic methods for ERM detection in OCT images is a relatively new research problem, and as such only a few studies on this subject have been published in recent years. The common denominator in the state-of-theart methods is the application of machine learning algorithms including multilayer perceptrons, random forests [16], [18], support vector machines and convolutional neural networks [17], [19], [20]. All these methods were developed and evaluated on private datasets, acquired with different OCT devices, annotated by a variable number of clinicial experts, and gathered following diverse exclusion criteria.…”

“…In contrast, there are few works concerned with automatic ERM detection. A literature review identified some research works that address this problem, [16]- [20]. A common factor to these studies is the use optical coherence tomography (OCT) retinal images as the source of information for the detection algorithm.…”

Automatic Segmentation and Intuitive Visualisation of the Epiretinal Membrane in 3D OCT Images Using Deep Convolutional Approaches

“…In [30], the authors proposed a methodology to identify the ILM layer and then analyse the neighbouring region and determine the presence or absence of the ERM disease using a classical machine learning strategy. In the work of Lo et al [31], the authors used a ResNet architecture for the screening of ERM in cross-sectional OCT images. Kuwayama et al [32] proposed a deep learning strategy with image augmentation for automated detection of different macular diseases, including ERM.…”