A Framework for the Discovery of Retinal Biomarkers in Optical Coherence Tomography Angiography (OCTA)

Abstract

Recent studies have demonstrated the potential of OCTA retinal imaging for the discovery of biomarkers of vascular disease of the eye and other organs. Furthermore, advances in deep learning have made it possible to train algorithms for the automated detection of such biomarkers. However, two key limitations of this approach are the need for large numbers of labeled images to train the algorithms, which are often not met by the typical single-centre prospective studies in the literature, and the lack of interpretability of the features learned during training. In the current study, we developed a network analysis framework to characterise retinal vasculature where geometric and topological information are exploited to increase the performance of classifiers trained on tens of OCTA images. We demonstrate our approach in two different diseases with a retinal vascular footprint: diabetic retinopathy (DR) and chronic kidney disease (CKD). Our approach enables the discovery of previously unreported retinal vascular morphological differences in DR and CKD, and demonstrate the potential of OCTA for automated disease assessment.

Euler characteristic surfaces

“…The images can be thought of as a grayscale images with an intensity value where higher values correspond to the presence of blood vessels. In many applications, these blood vessels are first extracted as a graph or tree and then certain graph features are used as input for classification algorithms [69,27].…”

“…Biomarkers. The predominant approaches for image analysis in the OCTA literature have centered around a small number of explainable candidate biomarkers as suggested by the clinical knowledge acquired [26]. Indeed, quantifiable features can be extracted from the OCTA images which are important biomarkers for DR. Statistical studies have identified the usefulness of local and global metrics based on the morphology of the foveal avascular zone (FAZ) and vascular-based metrics as biomarkers for distinguishing between healthy and DR eyes.…”

“…Indeed, quantifiable features can be extracted from the OCTA images which are important biomarkers for DR. Statistical studies have identified the usefulness of local and global metrics based on the morphology of the foveal avascular zone (FAZ) and vascular-based metrics as biomarkers for distinguishing between healthy and DR eyes. Examples of the former include FAZ area, FAZ contour irregularity [38,61], while examples of the latter are vessel caliber (VC), fractal dimension (FD), tortuosity, vessel density (VD) and geometric features of the vascular network [27,4,40,3,57]. For a more detailed review, please refer to [69].…”

Euler Characteristic Surfaces

“…The images can be thought of as a grayscale images with an intensity value where higher values correspond to the presence of blood vessels. In many applications, these blood vessels are first extracted as a graph or tree and then certain graph features are used as input for classification algorithms [40,41].…”

Topological Detection of Alzheimer’s Disease Using Betti Curves

“…Topological data analysis (TDA) has been used to study other static anatomical data (e.g. retina images [5,17]) as well as dynamic data such as functional MRI (fMRI) scans [27]. fMRI represents activity in different parts of the brain by detecting blood flow.…”