The segmentation of pathological fluid lesions in optical coherence tomography (OCT), including intraretinal fluid, subretinal fluid, and pigment epithelial detachment, is of great importance for the diagnosis and treatment of various eye diseases such as neovascular age-related macular degeneration and diabetic macular edema. Although significant progress has been achieved with the rapid development of fully convolutional neural networks (FCN) in recent years, some important issues remain unsolved. First, pathological fluid lesions in OCT show large variations in location, size, and shape, imposing challenges on the design of FCN architecture. Second, fluid lesions should be continuous regions without holes inside. But the current architectures lack the capability to preserve the shape prior information. In this study, we introduce an FCN architecture for the simultaneous segmentation of three types of pathological fluid lesions in OCT. First, attention gate and spatial pyramid pooling modules are employed to improve the ability of the network to extract multi-scale objects. Then, we introduce a novel curvature regularization term in the loss function to incorporate shape prior information. The proposed method was extensively evaluated on public and clinical datasets with significantly improved performance compared with the state-of-the-art methods.
This study was performed to test the repeatability and reproducibility of measurements of peri-papillary capillaries from four optical coherence tomography angiography (OCTA) devices. 109 healthy eyes were imaged with four OCTA devices (Spectralis, Optovue, Triton and Cirrus). A 3 × 3 mm scan pattern centered on the disc was repeated twice by each device. En face images of superficial capillary plexus were screened and processed for calculation. Vessel length density (VLD) was calculated on four equally divided parts of a ring between two concentric circles manually centered on the disc. General linear model (GLM) was used to test the impact of device and location on VLD. Intraclass correlation coefficient (ICC) of VLD between repeated scans was calculated. Of 218 acquisitions, 36%, 92%, 76% and 88% were eligible for analysis from Spectralis, Optovue, Triton and Cirrus, respectively. ICC was 0.94, 0.90, 0.84 and 0.87 for the four devices. GLM showed measurements significantly varied among devices (P < 0.001) and locations (P < 0.001). Pairwise comparison showed Triton = Spectralis >Optovue >Cirrus, and temporal = nasal >superior = inferior in measuring capillary VLD. This study revealed the repeatability of measuring peri-papillary capillaries was high for all four devices, while the reproducibility among the machines was unfavorable.
21 July 2022 in IEEE Transactions on Medical Imaging
Optical coherence tomography angiography (OCTA) is an imaging modality that can be used for analyzing retinal vasculature. Quantitative assessment of en face OCTA images requires accurate segmentation of the capillaries. Using deep learning approaches for this task faces two major challenges. First, acquiring sufficient manual delineations for training can take hundreds of hours. Second, OCTA images suffer from numerous contrast-related artifacts that are currently inherent to the modality and vary dramatically across scanners. We propose to solve both problems by learning a disentanglement of an anatomy component and a local contrast component from paired OCTA scans. With the contrast removed from the anatomy component, a deep learning model that takes the anatomy component as input can learn to segment vessels with a limited portion of the training images being manually labeled. Our method demonstrates state-of-the-art performance for OCTA vessel segmentation.
13 October 2022 in IEEE Transactions on Medical Imaging
Automatic segmentation and differentiation of retinal arteriole and venule (AV), defined as small blood vessels directly before and after the capillary plexus, are of great importance for the diagnosis of various eye diseases and systemic diseases, such as diabetic retinopathy, hypertension, and cardiovascular diseases. Optical coherence tomography angiography (OCTA) is a recent imaging modality that provides capillary-level blood flow information. However, OCTA does not have the colorimetric and geometric differences between AV as the fundus photography does. Various methods have been proposed to differentiate AV in OCTA, which typically needs the guidance of other imaging modalities. In this study, we propose a cascaded neural network to automatically segment and differentiate AV solely based on OCTA. A convolutional neural network (CNN) module is first applied to generate an initial segmentation, followed by a graph neural network (GNN) to improve the connectivity of the initial segmentation. Various CNN and GNN architectures are employed and compared. The proposed method is evaluated on multi-center clinical datasets, including 3×3 mm 2 and 6×6 mm 2 OCTA. The proposed method holds the potential to enrich OCTA image information for the diagnosis of various diseases.