Artificial intelligence-based prediction of diabetic retinopathy evolution (EviRed): protocol for a prospective cohort.

INTRODUCTION: An important obstacle in the fight against diabetic retinopathy (DR) is the use of a classification system based on old imaging techniques and insufficient data to accurately predict its evolution. New imaging techniques generate new valuable data, but we lack an adapted classification based on these data. The main objective of the Evaluation Intelligente de la Rétinopathie Diabétique, Intelligent evaluation of DR (EviRed) project is to develop and validate a system assisting the ophthalmologist in decision-making during DR follow-up by improving the prediction of its evolution. METHODS AND ANALYSIS: A cohort of up to 5000 patients with diabetes will be recruited from 18 diabetology departments and 14 ophthalmology departments, in public or private hospitals in France and followed for an average of 2 years. Each year, systemic health data as well as ophthalmological data will be collected. Both eyes will be imaged by using different imaging modalities including widefield photography, optical coherence tomography (OCT) and OCT-angiography. The EviRed cohort will be divided into two groups: one group will be randomly selected in each stratum during the inclusion period to be representative of the general diabetic population. Their data will be used for validating the algorithms (validation cohort). The data for the remaining patients (training cohort) will be used to train the algorithms. ETHICS AND DISSEMINATION: The study protocol was approved by the French South-West and Overseas Ethics Committee 4 on 28 August 2020 (CPP2020-07-060b/2020-A01725-34/20.06.16.41433). Prior to the start of the study, each patient will provide a written informed consent documenting his or her agreement to participate in the clinical trial. Results of this research will be disseminated in peer-reviewed publications and conference presentations. The database will also be available for further study or development that could benefit patients. TRIAL REGISTRATION NUMBER: NCT04624737.

Pseudoaveraging for denoising of OCT angiography: a deep learning approach for image quality enhancement in healthy and diabetic eyes.

BACKGROUND: This study aimed to develop a deep learning (DL) algorithm that enhances the quality of a single-frame enface OCTA scan to make it comparable to 4-frame averaged scan without the need for the repeated acquisitions required for averaging. METHODS: Each of the healthy eyes and eyes from diabetic subjects that were prospectively enrolled in this cross-sectional study underwent four repeated 6 × 6 mm macular scans (PLEX Elite 9000 SS-OCT), and the repeated scans of each eye were co-registered to produce 4-frame averages. This prospective dataset of original (single-frame) enface scans and their corresponding averaged scans was divided into a training dataset and a validation dataset. In the training dataset, a DL algorithm (named pseudoaveraging) was trained using original scans as input and 4-frame averages as target. In the validation dataset, the pseudoaveraging algorithm was applied to single-frame scans to produce pseudoaveraged scans, and the single-frame and its corresponding averaged and pseudoaveraged scans were all qualitatively compared. In a separate retrospectively collected dataset of single-frame scans from eyes of diabetic subjects, the DL algorithm was applied, and the produced pseudoaveraged scan was qualitatively compared against its corresponding original. RESULTS: This study included 39 eyes that comprised the prospective dataset (split into 5 eyes for training and 34 eyes for validating the DL algorithm), and 105 eyes that comprised the retrospective test dataset. Of the total 144 study eyes, 58% had any level of diabetic retinopathy (with and without diabetic macular edema), and the rest were from healthy eyes or eyes of diabetic subjects but without diabetic retinopathy and without macular edema. Grading results in the validation dataset showed that the pseudoaveraged enface scan ranked best in overall scan quality, background noise reduction, and visibility of microaneurysms (p < 0.05). Averaged scan ranked best for motion artifact reduction (p < 0.05). Grading results in the test dataset showed that pseudoaveraging resulted in enhanced small vessels, reduction of background noise, and motion artifact in 100%, 82%, and 98% of scans, respectively. Rates of false-positive/-negative perfusion were zero. CONCLUSION: Pseudoaveraging is a feasible DL approach to more efficiently improve enface OCTA scan quality without introducing notable image artifacts.

Hybrid Fusion of High-Resolution and Ultra-Widefield OCTA Acquisitions for the Automatic Diagnosis of Diabetic Retinopathy.

Optical coherence tomography angiography (OCTA) can deliver enhanced diagnosis for diabetic retinopathy (DR). This study evaluated a deep learning (DL) algorithm for automatic DR severity assessment using high-resolution and ultra-widefield (UWF) OCTA. Diabetic patients were examined with 6×6 mm2 high-resolution OCTA and 15×15 mm2 UWF-OCTA using PLEX®Elite 9000. A novel DL algorithm was trained for automatic DR severity inference using both OCTA acquisitions. The algorithm employed a unique hybrid fusion framework, integrating structural and flow information from both acquisitions. It was trained on data from 875 eyes of 444 patients. Tested on 53 patients (97 eyes), the algorithm achieved a good area under the receiver operating characteristic curve (AUC) for detecting DR (0.8868), moderate non-proliferative DR (0.8276), severe non-proliferative DR (0.8376), and proliferative/treated DR (0.9070). These results significantly outperformed detection with the 6×6 mm2 (AUC = 0.8462, 0.7793, 0.7889, and 0.8104, respectively) or 15×15 mm2 (AUC = 0.8251, 0.7745, 0.7967, and 0.8786, respectively) acquisitions alone. Thus, combining high-resolution and UWF-OCTA acquisitions holds the potential for improved early and late-stage DR detection, offering a foundation for enhancing DR management and a clear path for future works involving expanded datasets and integrating additional imaging modalities.

Retinal capillary and choriocapillaris assessment using a beam modifier optical coherence tomography angiography module to increase lateral optical resolution.

PURPOSE: To assess a new optical coherence tomography angiography (OCTA) technology and its contribution to retinal vascularization and choriocapillaris (CC) exploration. METHODS: A new module, named "Beam expander" (BE), which increases the lateral resolution of OCTA, was used in combination with a prototype software in the PLEX® Elite 9000 Swept-Source OCT instrument (ZEISS, Dublin, CA). This prospective study involved 22 healthy subjects imaged with and without BE. Qualitative analysis of superficial capillary plexus (SCP), deep capillary complex (DCC) retinal and CC angiograms were performed. Perfusion density (PD), vessel density (VD), and foveal avascular zone (FAZ) measurements were also compared. RESULTS: Qualitative analysis of single SCP and DCC retinal angiograms acquired with BE showed significantly better vessel sharpness (respectively, p = 0.0002, and p<0.0001), and greater peripheral image quality (p = 0.028 and p = 0.007) compared to standard OCTA images. Mean VD of whole retina single scans was significantly higher for BE angiograms compared to classic angiograms (28.16 ±1.29 mm-1 and 23.36 ±0.92 mm-1, respectively, p<0.0001). Repeatability of VD, PD and FAZ raw size were found to be similar between the two methods (intraclass correlation coefficient: 0.671, 0.604 and 0.994 with BE versus 0.764, 0.638 and 0.990 without BE). CC image quality was found to be significantly superior with BE, and flow deficits were more visible in all BE scans compared to standard scans. CONCLUSIONS: An increase in lateral resolution of the OCT beam resulted in higher quality of retinal and choriocapillaris OCTA images in healthy subjects. These results provide significant insights into the future OCTA imaging enhancements.

Optical Coherence Tomography Angiography Macular and Peripapillary Vessel Perfusion Density in Healthy Subjects, Glaucoma Suspects, and Glaucoma Patients.

Purpose: To evaluate macular and peripapillary vessel perfusion density (VD) in glaucoma suspects (GS) and glaucoma patients; to correlate ganglion cell-inner plexiform layer (GCIPL) and retinal nerve fiber layer (RNFL) thicknesses with macular and peripapillary VD; and to evaluate the diagnostic accuracy of the structural and vascular parameters. Methods: A consecutive series of GS, glaucoma patients, and healthy subjects was prospectively recruited from July 1, 2016, to January 31, 2017. All subjects underwent standard automated perimetry, spectral-domain optical coherence tomography (OCT), and 6 × 6-mm optical coherence tomography angiography (OCT-A) centered on the fovea and optic nerve. Results: Forty controls, 40 GS, and 40 glaucoma patients were enrolled. Peripapillary RNFL, GCIPL, and macular RNFL thicknesses significantly decreased in the glaucoma group compared to controls and GS (P < 0.01). Peripapillary VD in average and in the superior and inferior quadrants decreased in the glaucoma group (P ≤ 0.001); conversely, macular VD was not statistically different across groups (P > 0.05). At the peripapillary area, a correlation between RNFL thickness and VD was found; conversely, no statistically significant correlation was found between GCIPL thicknesses and macular VD (all P > 0.05) in all groups. Peripapillary RNFL and GCIPL showed higher diagnostic capacity compared to peripapillary and macular VDs. Conclusions: Structural damage is evident both in the peripapillary and in macular areas. Vascular damage seems to be less prominent, as it was seen only for the glaucoma group and at the radial peripapillary plexus. Diagnostic abilities are excellent for structural variables, less so but still good for peripapillary VD, and poor for macular VD.

Predictive Value of Outer Retina En Face OCT Imaging for Geographic Atrophy Progression.

PURPOSE: We determined if the ellipsoid zone (EZ) disruption pattern could be predictive of the geographic atrophy (GA) pattern at 1 year in dry age-related macular degeneration (AMD). METHODS: A retrospective study was done of dry eyes in patients with AMD and GA from July to November 2013. Eyes with previous choroidal neovascularization were excluded. Based on spectral domain optical coherence tomography (SD-OCT), the GA was assessed at each timepoint, using a sub-RPE slab derived from the Cirrus Advanced RPE Analysis software encompassing the RPE (sub-RPE slab). Disruption of the EZ also was assessed at baseline, using en face extraction of a 20-µm-thick slab, 20 µm above the RPE (EZ slab) encompassing the EZ band using two different algorithms (RPE and RPE-fit). The EZ disruption area surrounding GA at baseline was quantified using ImageJ software. Primary endpoint was to identify en face pattern similarities between the baseline EZ disruption and the 1-year GA. Secondary endpoint was to correlate the baseline EZ disruption area surrounding GA with the GA enlargement over 1 year. Statistical analysis was performed using a correlation test (Pearson) and a t-test. RESULTS: We included 37 eyes of 31 patients with dry AMD. En face EZ disruption pattern correlated in two-thirds of cases with the 1-year GA pattern using both algorithms. The EZ disruption area surrounding GA at baseline and GA enlargement over 1 year were poorly correlated when RPE-fit algorithm (R = 0.17) was used. The correlation was still poor using an RPE algorithm (R = 0.38), but increased after selection of eyes without reticular pseudodrusen (R = 0.79). CONCLUSIONS: The EZ disruption pattern could be an indicator for GA pattern progression, but is not a good quantitative tool to predict the size of GA in the overall population over a 1-year period except for patients without reticular pseudodrusen. The results in this specific population must be confirmed by further studies.