Ultra-Widefield OCT Angiography

Abstract

Optical Coherence Tomography Angiography (OCTA), a functional extension of OCT, has the potential to replace most invasive fluorescein angiography (FA) exams in ophthalmology. So far, OCTA's field of view is however still lacking behind fluorescence fundus photography techniques. This is problematic, because many retinal diseases manifest at an early stage by changes of the peripheral retinal capillary network. It is therefore desirable to expand OCTA's field of view to match that of ultrawidefield fundus cameras. We present a custom developed clinical high-speed swept-source OCT (SS-OCT) system operating at an acquisition rate 8-16 times faster than today's state-of-the-art commercially available OCTA devices. Its speed allows us to capture ultra-wide fields of view of up to 90 degrees with an unprecedented sampling density and hence extraordinary resolution by merging two single shot scans with 60 degrees in diameter. To further enhance the visual appearance of the angiograms, we developed for the first time a three-dimensional deep learning based algorithm for denoising volumetric OCTA data sets. We showcase its imaging performance and clinical usability by presenting images of patients suffering from diabetic retinopathy.

Stable classification of diabetic structures from incorrectly labeled optical coherence tomography angiography en face images using multi instance learning

“…Data acquired for this study utilized a scan protocol that imaged over a field of view (FoV) of 65°(18mm x 18mm) with two repetitions per B-scan location, sampled at 2048 x 2044 samples laterally. 3 MIL is a technique which has been applied in combination with ML-based classifiers for the classification of DR in fundus photography (FP) in previously published work. In the context of ML MIL treats data, a so-called bag, which is either positive of negative, as a container for a manifold of smaller feature spaces, coming from that bag, so called instances.…”

“…Widefield OCTA scans provide information of the peripheral retinal areas, while the latest fast acquisition times allow for comfortable imaging of the patients. 3,4 artificial intelligence-algorithms have been utilized in the past to classify DR and even its severity. Most algorithms, however, require precise labels indicating obvious or known lesions.…”

Stable classification of diabetic structures from incorrectly labeled OCTA en face images using multi instance learning

“…OCTA was calculated as the variance of every voxel and processed in a batch-like manner. 5 Our proposed architecture, MIL-ResNet14 (MIL-ResNet14), is an instance-based learning technique since we assume that information for classification of DR is held on an instance level. For pre-processing we take the entire en face image, normalize (set pixel value mean µ = 0 with a standard deviation σ = 1) and denoise it (apply Gaussian filter with σ = 1).…”

“…Recent developments in laser technology allow fast and comfortable scanning of large field of views (FoVs) in patients. 5 This technique is often referred to as widefield OCTA (Widefield OCTA). Because Widefield OCTA datasets are very large, physicians focus diagnosis on a limited number of carefully selected cross-sectional scans, called B-scans, or projection maps of the retinal slab along the optical axis, often called en face images.…”

Characteristics and Classification of Choroidal Caverns in Patients with Various Retinal and Chorioretinal Diseases

“…The development of the optical coherence tomography (OCT) imaging technology, especially the enhanced depth imaging [ 1 , 2 ] and swept-source OCT (SS-OCT) [ 3 , 4 , 5 ], has made it possible to detect the fine structure of choroid. Choroidal caverns, small lesions within the choroid, were first discovered in 2016 by Querques et al [ 6 ] in patients with geographic atrophy (GA).…”