Search

Abstract Quantitative assessment of retinal microvasculature in optical coherence tomography angiography (OCTA) images is important for studying, diagnosing, monitoring, and guiding the treatment of ocular and systemic diseases. However, the OCTA user community lacks universal and transparent image analysis tools that can be applied to images from a range of OCTA instruments and provide reliable and consistent microvascular metrics from diverse datasets. We present a retinal extension to the OCTA Vascular Analyser (OCTAVA) that addresses the challenges of providing robust, easy-to-use, and transparent analysis of retinal OCTA images. OCTAVA is a user-friendly, open-source toolbox that can analyse retinal OCTA images from various instruments. The toolbox delivers seven microvascular metrics for the whole image or subregions and six metrics characterising the foveal avascular zone. We validate OCTAVA using images collected by four commercial OCTA instruments demonstrating robust performance across datasets from different instruments acquired at different sites from different study cohorts. We show that OCTAVA delivers values for retinal microvascular metrics comparable to the literature and reduces their variation between studies compared to their commercial equivalents. By making OCTAVA publicly available, we aim to expand standardised research and thereby improve the reproducibility of quantitative analysis of retinal microvascular imaging. Such improvements will help to better identify more reliable and sensitive biomarkers of ocular and systemic diseases.

Abstract Deep learning methods have enabled a fast, accurate and automated approach for retinal layer segmentation in posterior segment OCT images. Due to the success of semantic segmentation methods adopting the U-Net, a wide range of variants and improvements have been developed and applied to OCT segmentation. Unfortunately, the relative performance of these methods is difficult to ascertain for OCT retinal layer segmentation due to a lack of comprehensive comparative studies, and a lack of proper matching between networks in previous comparisons, as well as the use of different OCT datasets between studies. In this paper, a detailed and unbiased comparison is performed between eight U-Net architecture variants across four different OCT datasets from a range of different populations, ocular pathologies, acquisition parameters, instruments and segmentation tasks. The U-Net architecture variants evaluated include some which have not been previously explored for OCT segmentation. Using the Dice coefficient to evaluate segmentation performance, minimal differences were noted between most of the tested architectures across the four datasets. Using an extra convolutional layer per pooling block gave a small improvement in segmentation performance for all architectures across all four datasets. This finding highlights the importance of careful architecture comparison (e.g. ensuring networks are matched using an equivalent number of layers) to obtain a true and unbiased performance assessment of fully semantic models. Overall, this study demonstrates that the vanilla U-Net is sufficient for OCT retinal layer segmentation and that state-of-the-art methods and other architectural changes are potentially unnecessary for this particular task, especially given the associated increased complexity and slower speed for the marginal performance gains observed. Given the U-Net model and its variants represent one of the most commonly applied image segmentation methods, the consistent findings across several datasets here are likely to translate to many other OCT datasets and studies. This will provide significant value by saving time and cost in experimentation and model development as well as reduced inference time in practice by selecting simpler models.

Abstract Adaptive optics flood illumination ophthalmoscopy (AO-FIO) is an established imaging tool in the investigation of retinal diseases. However, the clinical interpretation of AO-FIO images can be challenging due to varied image quality. Therefore, image quality assessment is essential before interpretation. An image assessment tool will also assist further work on improving the image quality, either during acquisition or post processing. In this paper, we describe, validate and compare two automated image quality assessment methods; the energy of Laplacian focus operator (LAPE; not commonly used but easily implemented) and convolutional neural network (CNN; effective but more complex approach). We also evaluate the effects of subject age, axial length, refractive error, fixation stability, disease status and retinal location on AO-FIO image quality. Based on analysis of 10,250 images of 50 × 50 μm size, at 41 retinal locations, from 50 subjects we demonstrate that CNN slightly outperforms LAPE in image quality assessment. CNN achieves accuracy of 89%, whereas LAPE metric achieves 73% and 80% (for a linear regression and random forest multiclass classifier methods, respectively) compared to ground truth. Furthermore, the retinal location, age and disease are factors that can influence the likelihood of poor image quality.

Abstract Inflammation and cigarette smoking predispose to macular diseases, and choroidal and retinal thinning. We explored the choroidal and retinal thicknesses in young adults against their 7-year C-reactive protein (CRP) level trajectory and pack-years smoked. Participants from the Raine study, a longitudinal cohort study, had serum CRP levels analysed at the 14-, 17-, and 20-year follow-ups. Group-based trajectory modelling was used to classify participants according to their 7-year CRP levels. At the 20-year follow-up (at 18–22 years old), participants completed questionnaires on their smoking history, and underwent optical coherence tomography imaging to obtain their choroidal and retinal thicknesses at the macula. Three CRP trajectories were identified: consistently low CRP levels (78% of sample), increasing (11%), or consistently high (11%). 340 and 1035 participants were included in the choroidal and retinal thickness analyses, respectively. Compared to those in the “Low” trajectory group, participants in the “Increasing” and “High” groups had 14–21 μm thinner choroids at most macular regions. Every additional pack-year smoked was linked with a 0.06–0.10 μm thinner retina at the inner and outer macular rings, suggesting a dose-dependent relationship between smoking and thinner retinas. These associations may suggest that an increased risk of future visual impairment or eye disease associated with these risk factors may be present since young adulthood.