A Multitask Deep-Learning System to Classify Diabetic Macular Edema for Different Optical Coherence Tomography Devices: A Multicenter Analysis

Abstract

<a><b>Objective:</b></a> Diabetic macular edema (DME) is the primary cause of vision loss among individuals with diabetes mellitus (DM). We developed, validated, and tested a deep-learning (DL) system for classifying DME using images from three common commercially available optical coherence tomography (OCT) devices. <p><b>Research Design and Methods:</b> We trained and validated two versions of a multi-task convolution neural network (CNN) to classify DME (center-involved DME [CI-DME], non-CI-DME, or absence of DME) using three-dimensional (3D) volume-scans and two-dimensional (2D) B-scans respectively. For both 3D and 2D CNNs, we employed the residual network (ResNet) as the backbone. For the 3D CNN, we used a 3D version of ResNet-34 with the last fully connected layer removed as the feature extraction module. A total of 73,746 OCT images were used for training and primary validation. External testing was performed using 26,981 images across seven independent datasets from Singapore, Hong Kong, the US, China, and Australia. </p> <p><b>Results:</b> In classifying the presence or absence of DME, the DL system achieved area under the receiver operating characteristic curves (AUROCs) of 0.937 (95% CI 0.920–0.954), 0.958 (0.930–0.977), and 0.965 (0.948–0.977) for primary dataset obtained from Cirrus, Spectralis, and Triton OCTs respectively, in addition to AUROCs greater than 0.906 for the external datasets. For the further classification of the CI-DME and non-CI-DME subgroups, the AUROCs were 0.968 (0.940–0.995), 0.951 (0.898–0.982), and 0.975 (0.947–0.991) for the primary dataset and greater than 0.894 for the external datasets. </p> <p><b>Conclusion:</b> We demonstrated excellent performance with a DL system for the automated classification of DME, highlighting its potential as a promising second-line screening tool for patients with DM, which may potentially create a more effective triaging mechanism to eye clinics. </p>

Optical Coherence Tomography Classification Systems for Diabetic Macular Edema and Their Associations With Visual Outcome and Treatment Responses – An Updated Review

“…Another DL model trained with data from SDOCT was found to detect center-involved DME on color fundus photographs with similar sensitivity but better specificity, compared with retinal specialists 95 . Recently, a multitask DL system has been trained and validated to classify any DME, center-involved DME, non–center-involved DME, and non-DME retinal abnormalities from multiple devices, and the results are promising 96 . The recognition of morphological patterns and detection of retinal fluid with DL has achieved a high accuracy and is also found to be indistinguishable from clinicians 97,98 .…”

“…95 Recently, a multitask DL system has been trained and validated to classify any DME, centerinvolved DME, non-center-involved DME, and non-DME retinal abnormalities from multiple devices, and the results are promising. 96 The recognition of morphological patterns and detection of retinal fluid with DL has achieved a high accuracy and is also found to be indistinguishable from clinicians. 97,98 Attempts have been made to correlate volumetric change of retinal fluid during anti-VEGF treatments of DME using DL with functional outcomes.…”

Latent Space Representational Learning of Deep Features for Acute Lymphoblastic Leukemia Diagnosis

“…Deep Learning-based computer-aided diagnosis systems have been recognized of their efficiency in automatically identifying the existence of many diseases [15] such as diabetic retinopathy [16], several kinds of cancers [14], and Covid-19 infection [17][18][19]. In computer vision applications, the architecture of CAD systems is specifically based on convolutional neural networks, or convnets, which are utilized for image segmentation [18], classification, object detection and recognition [14,[20][21][22][23][24]. Although convnets offer higher classification performance than traditional classifiers, the training of a custom CNN that is built from scratch is computationally intensive, time consuming and requires large datasets to provide acceptable performance [25].…”