Strong versus Weak Data Labeling for Artificial Intelligence Algorithms in the Measurement of Geographic Atrophy.

Purpose: To gain an understanding of data labeling requirements to train deep learning models for measurement of geographic atrophy (GA) with fundus autofluorescence (FAF) images. Design: Evaluation of artificial intelligence (AI) algorithms. Subjects: The Age-Related Eye Disease Study 2 (AREDS2) images were used for training and cross-validation, and GA clinical trial images were used for testing. Methods: Training data consisted of 2 sets of FAF images; 1 with area measurements only and no indication of GA location (Weakly labeled) and the second with GA segmentation masks (Strongly labeled). Main Outcome Measures: Bland-Altman plots and scatter plots were used to compare GA area measurement between ground truth and AI measurements. The Dice coefficient was used to compare accuracy of segmentation of the Strong model. Results: In the cross-validation AREDS2 data set (n = 601), the mean (standard deviation [SD]) area of GA measured by human grader, Weakly labeled AI model, and Strongly labeled AI model was 6.65 (6.3) mm2, 6.83 (6.29) mm2, and 6.58 (6.24) mm2, respectively. The mean difference between ground truth and AI was 0.18 mm2 (95% confidence interval, [CI], -7.57 to 7.92) for the Weakly labeled model and -0.07 mm2 (95% CI, -1.61 to 1.47) for the Strongly labeled model. With GlaxoSmithKline testing data (n = 156), the mean (SD) GA area was 9.79 (5.6) mm2, 8.82 (4.61) mm2, and 9.55 (5.66) mm2 for human grader, Strongly labeled AI model, and Weakly labeled AI model, respectively. The mean difference between ground truth and AI for the 2 models was -0.97 mm2 (95% CI, -4.36 to 2.41) and -0.24 mm2 (95% CI, -4.98 to 4.49), respectively. The Dice coefficient was 0.99 for intergrader agreement, 0.89 for the cross-validation data, and 0.92 for the testing data. Conclusions: Deep learning models can achieve reasonable accuracy even with Weakly labeled data. Training methods that integrate large volumes of Weakly labeled images with small number of Strongly labeled images offer a promising solution to overcome the burden of cost and time for data labeling. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Assessing the Sensitivity of OCT-A Retinal Vasculature Metrics.

Purpose: The purpose of this study was to examine the sensitivity of quantitative metrics of the retinal vasculature derived from optical coherence tomography angiography (OCT-A) images. Methods: Full retinal vascular slab OCT-A images were obtained from 94 healthy participants. Capillary loss, at 1% increments up to 50%, was simulated by randomly removing capillary segments (1000 iterations of randomized loss for each participant at each percent loss). Thirteen quantitative metrics were calculated for each image: foveal avascular zone (FAZ) area, vessel density, vessel complexity index (VCI), vessel perimeter index (VPI), fractal dimension (FD), and parafoveal intercapillary area (PICA) measurements with and without the FAZ (mean PICA, summed PICA, PICA regularity, and PICA standard deviation [PICA SD]). The sensitivity of each metric was calculated as the percent loss at which 80% of the iterations for a participant fell outside of two standard deviations from the sample's normative mean. Results: The most used OCT-A metrics, FAZ area and vessel density, were not significantly different from normative values until 27.69% and 16.00% capillary loss, respectively. Across the remaining metrics, metric sensitivity ranged from 6.37% (PICA SD without FAZ) to 39.78% (Summed PICA without FAZ). Conclusions: The sensitivity of vasculature metrics for detecting random capillary loss varies substantially. Further efforts simulating different patterns of capillary loss are needed for comparison. Additionally, mapping the repeatability of metrics over time in a normal population is needed to further define metric sensitivity. Translational Relevance: Quantitative metrics vary in their ability to detect vascular abnormalities in OCT-A images. Metric choice in screening studies will need to balance expected capillary abnormalities and the quality of the OCT-A images being used.

The Shape of the Foveal Avascular Zone: When a Circle Isn't Round.

Translational Relevance: Quantitative assessment of OCT-A images includes evaluating circularity and roundness of the FAZ. Inconsistent or inaccurate mathematical definitions of these metrics impacts their utility as biomarkers and impairs the ability to combine and compare results across studies.

The Impact of Axial Eye Growth on Foveal Avascular Zone Measurements in Children.

SIGNIFICANCE: Foveal avascular zone (FAZ) area is a frequently used biomarker in diseases impacting the retinal vasculature in pediatric populations. Variation in axial length between individuals results in differences in lateral image scale, which affect the accuracy of FAZ area measurements. Accordingly, changes in axial length over time within individual children would affect estimates of FAZ area change. PURPOSE: This study aimed to quantify how changes in axial length over time affect estimates of FAZ area change using optical coherence tomography angiography (OCT-A) images. METHODS: Twenty pediatric participants (<18 years old) and 40 adult participants were imaged on Optovue's Avanti system (Fremont, CA) and had axial length measurements acquired at two time points. The FAZ was segmented twice using the OCT-A image at each time point. Foveal avascular zone area was estimated at both time points using the assumed/fixed axial length of the OCT-A device (unscaled) and using the participant's axial length (scaled). Changes in FAZ area over time were compared between the pediatric and adult groups using both unscaled and scaled data. RESULTS: The average ± standard deviation follow-up time was 3.35 ± 1.66 years for the pediatric group and 2.90 ± 1.65 years for the adult group. Using unscaled data, FAZ area seemed to decrease between visits in the pediatric group (P = .004), whereas the FAZ area increased between visits in the adult group (P = .003). When correctly scaled data were used, the FAZ area still increased between visits for the adult group (P < .001), although the FAZ area no longer showed a significant change between visits for the pediatric group (P = .37). When comparing the normalized FAZ area change across visits between unscaled and scaled data, a significant difference was found between the adult and pediatric groups (P < .001). CONCLUSIONS: Scaled data should be used when measuring FAZ area in pediatric populations, especially in longitudinal studies.

Retinal alterations in patients with Lafora disease.

PURPOSE: Lafora disease is a genetic neurodegenerative metabolic disorder caused by insoluble polyglucosan aggregate accumulation throughout the central nervous system and body. The retina is an accessible neural tissue, which may offer alternative methods to assess neurological diseases quickly and noninvasively. In this way, noninvasive imaging may provide a means to characterize neurodegenerative disease, which enables earlier identification and diagnosis of disease and the ability to monitor disease progression. In this study, we sought to characterize the retina of individuals with Lafora disease using non-invasive retinal imaging. METHODS: One eye of three individuals with genetically confirmed Lafora disease were imaged with optical coherence tomography (OCT) and adaptive optics scanning light ophthalmoscopy (AOSLO). When possible, OCT volume and line scans were acquired to assess total retinal thickness, ganglion cell-inner plexiform layer thickness, and outer nuclear layer + Henle fiber layer thickness. OCT angiography (OCTA) scans were acquired in one subject at the macula and optic nerve head (ONH). AOSLO was used to characterize the photoreceptor mosaic and examine the retinal nerve fiber layer (RNFL). RESULTS: Two subjects with previous seizure activity demonstrated reduced retinal thickness, while one subject with no apparent symptoms had normal retinal thickness. All other clinical measures, as well as parafoveal cone density, were within normal range. Nummular reflectivity at the level of the RNFL was observed using AOSLO in the macula and near the ONH in all three subjects. CONCLUSIONS: This multimodal retinal imaging approach allowed us to observe a number of retinal structural features in all three individuals. Most notably, AOSLO revealed nummular reflectivity within the inner retina of each subject. This phenotype has not been reported previously and may represent a characteristic change produced by the neurodegenerative process.

Assessing Foveal Structure in Individuals with TYR R402Q and S192Y Hypomorphic Alleles.

Purpose: To assess the impact of two TYR hypomorphic alleles (R402Q and S192Y) on foveal pit and foveal avascular zone (FAZ) morphology. Design: Prospective, cross-sectional study. Participants: A total of 164 participants with normal vision (67 male and 97 female; mean ± standard deviation [SD] age = 30.5 ± 12.8 years) were recruited. Methods: Sequencing of more than 100 pigmentation-related genes was performed, and results were reviewed for the presence or absence of the TYR polymorphisms R402Q (rs1126809) and S192Y (rs1042602). Volumetric scans of the macula were obtained for each participant using OCT, and retinal thickness maps were analyzed using custom software. OCT angiography was used to image the FAZ, which was manually segmented and measured. Linear mixed model analysis was used to assess associations between genotype and foveal pit morphology. Main Outcome Measures: Foveal pit depth, diameter, volume, and FAZ area in relation to the presence of hypomorphic alleles R402Q and S192Y on the TYR gene. Results: Heterozygosity for the TYR R402Q allele was associated with decreased pit diameter (P = 0.0094) and decreased FAZ area (P = 0.025). Homozygosity for the TYR R402Q allele was associated with reduced pit volume (P = 0.0005), decreased pit depth (P = 0.007), reduced pit diameter (P = 0.0052), and reduced FAZ area (P = 0.0012). Homozygosity for TYR S192Y was associated with reduced FAZ area (P = 0.016). Heterozygosity for the TYR S192Y allele was not associated with differences in foveal pit depth, diameter, volume, or FAZ area (P > 0.05). Conclusions: Although the role of the TYR R402Q and S192Y hypomorphic alleles in albinism remains controversial, our data suggest that these variants contribute to the extensive inter-individual variability in foveal morphology in the normal population. Our results contribute to the evolving picture of the relationship between ocular pigmentation and foveal morphology.

Assessing Interocular Symmetry of the Foveal Cone Mosaic.

Purpose: To test the hypothesis that foveal cone topography is symmetrical between contralateral eyes. Methods: We used adaptive optics scanning light ophthalmoscopy to acquire images of the foveal cone mosaic in each eye of 58 subjects with normal vision (35 female, 23 male). Cones were semiautomatically identified over a 300 × 300-µm foveal area. From these cone coordinates, maps of cone density were derived, and we extracted estimates of peak cone density from each map. Mosaic regularity was assessed using Voronoi cell area regularity (VCAR). Average roundness and average area of the 70%, 75%, 80%, 85%, and 90% of peak density isodensity contours were evaluated. Results: The average peak cone density for right eyes was 180,286 cones/mm2 (n = 49) and for left eyes was 182,397 cones/mm2 (n = 45), with a mean absolute difference of 6363 cones/mm2 (n = 43). Peak density, cone spacing, VCAR, and average area within the isodensity contours of fellow eyes were not significantly different (P = 0.60, P = 0.83, P = 0.30, and P = 0.39, respectively). However, the average roundness of the isodensity contours was 2% more circular in the right eyes than in the left eyes (P = 0.02). Conclusions: There is interocular symmetry of peak foveal cone density, mosaic regularity, and area encompassing the most densely packed cells in subjects with normal vision. The origin and significance of the observed interocular difference in average roundness of the isodensity contours are unclear.

Assessing the Influence of OCT-A Device and Scan Size on Retinal Vascular Metrics.

Purpose: The purpose of this study was to investigate the effect of device and scan size on quantitative optical coherence tomography angiography (OCT-A) metrics. Methods: The 3 × 3 mm scans from Optovue AngioVue and Zeiss AngioPlex systems were included for 18 eyes of 18 subjects without ocular pathology. The foveal avascular zone (FAZ) was segmented manually by two observers, from which estimates of FAZ area (using both the nominal image scale and the axial length corrected image scale) and acircularity were derived. Three scan sizes (3 mm, 6 mm HD, and 8 mm) from the AngioVue system were included for 15 eyes of 15 subjects without ocular pathology. For each subject, larger image sizes were resized to the same resolution as 3 × 3 mm scans, aligned, then cropped to a common area. FAZ area, FAZ acircularity, average and total parafoveal intercapillary area, vessel density, and vessel end points were computed. Results: Between the devices used here, there were no significant differences in FAZ acircularity (P = 0.88) or FAZ area using scaled (P = 0.11) or unscaled images (P = 0.069). Although there was no significant difference in FAZ area across scan sizes (P = 0.30), vessel morphometry metrics were all significantly influenced by scan size. Conclusions: The scan devices and sizes used here do not affect FAZ area measures derived from manual segmentations. In contrast, vessel morphometry metrics are affected by scan size. As individual differences in axial length induce differences in absolute scan size, extreme care should be taken when interpreting metrics of vessel morphometry, both between and within OCT-A devices. Translational Relevance: A better characterization of the confounds surrounding OCT-A retinal vasculature metrics can lead to improved application of these metrics as biomarkers for retinal and systemic diseases.

Preservation of the Foveal Avascular Zone in Achromatopsia Despite the Absence of a Fully Formed Pit.

Purpose: To examine the foveal avascular zone (FAZ) in patients with congenital achromatopsia (ACHM). Methods: Forty-two patients with genetically confirmed ACHM were imaged either with Optovue's AngioVue system or Zeiss's Plex Elite 9000, and the presence or absence of a FAZ was determined. For images where a FAZ was present and could be confidently segmented, FAZ area, circularity index, and roundness were measured and compared with previously published normative values. Structural optical coherence tomography images were acquired to assess the degree of foveal hypoplasia (number and thickness of inner retinal layers present at the fovea). Results: A FAZ was present in 31 of 42 patients imaged (74%), although no determination could be made for 11 patients due to poor image quality (26%). The mean ± SD FAZ area for the ACHM retina was 0.281 ± 0.112 mm2, which was not significantly different from the previously published normative values (P = 0.94). However, their FAZs had decreased circularity (P < 0.0001) and decreased roundness (P < 0.0001) compared to the normative cohort. In the patients with ACHM examined here, the FAZ area decreased as the number and thickness of the retained inner retinal layers increased. Conclusions: Our data demonstrate that despite the presence of foveal hypoplasia, patients with ACHM can have a FAZ. This is distinct from other conditions associated with foveal hypoplasia, which generally show an absence of the FAZ. In ACHM, FAZ formation does not appear to be sufficient for complete pit formation, contrary to some models of foveal development.

Interocular asymmetry of foveal avascular zone morphology and parafoveal capillary density in sickle cell retinopathy.

OBJECTIVES: To examine interocular asymmetry of foveal avascular zone (FAZ) and parafoveal capillary density metrics in sickle cell retinopathy (SCR) using optical coherence tomography angiography (OCT-A). METHODS: This cross-sectional, retrospective study evaluated SCR patients and unaffected controls who underwent 3x3mm macular OCT-A imaging using a spectral domain-OCT system. FAZ (area, perimeter, and acircularity index) and parafoveal capillary density metrics were computed for both eyes of each participant. In unaffected controls, interocular difference in FAZ and parafoveal capillary density metrics were evaluated using Bland-Altman plots. SCR patients with interocular difference outside the upper 97.5% and lower 2.5% limits of agreement from controls were defined as having interocular asymmetry. Area under receiver operating characteristic curve (AROC) was also performed to determine the ability of the absolute interocular difference to differentiate between subjects with SCR-including non-proliferative SCR (NP-SCR) and proliferative SCR (P-SCR)-and unaffected controls. RESULTS: Thirty-one patients with SCR (21 NP-SCR and 10 P-SCR) and 14 race-matched and age-matched controls were included for analysis. Interocular asymmetry was seen for all FAZ and parafoveal capillary density metrics in NP-SCR and P-SCR subjects. SCR subjects showed greater disease severity in the left-eye for FAZ and parafoveal capillary density metrics. CONCLUSIONS: NP-SCR and P-SCR patients demonstrated quantifiable interocular asymmetry in FAZ and parafoveal capillary density metrics compared to unaffected subjects, with left-eye predominance in disease severity.

Imaging of Macrophage-Like Cells in Living Human Retina Using Clinical OCT.

Purpose: To image retinal macrophages at the vitreoretinal interface in the living human retina using a clinical optical coherence tomography (OCT) device. Methods: Eighteen healthy controls and three patients with retinopathies were imaged using a clinical spectral-domain OCT. In controls, 10 sequential scans were collected at three different locations: (1) ∼9 degrees temporal to the fovea, (2) the macula, and (3) the optic nerve head (ONH). Intervisit repeatability was evaluated by imaging the temporal retina twice on the same day and 3 days later. Only 10 scans at the temporal retina were obtained from each patient. A 3-µm OCT reflectance (OCT-R) slab located above the inner limiting membrane (ILM) surface was averaged. Results: In controls, ramified macrophage-like cells with regular spatial separation were visualized in the temporal and ONH OCT-R images; however, cell structures were not resolvable at the macula. Interim changes in cell position suggestive of cell translocation were observed between images collected on the same day and those collected 3 days later. There was considerable variation in cell density and nearest-neighbor distance (NND) across controls. Mean ± SD cell densities measured at the temporal and ONH were 78 ± 23 cells/mm2 and 57 ± 16 cells/mm2, respectively. Similarly, mean ± SD NNDs measured at the temporal and ONH were 74.3 ± 13.3 µm and 93.3 ± 20.0 µm, respectively. Nonuniform spatial distribution and altered morphology of the cells were identified in patients with retinopathies. Conclusions: Our findings showed regular spatial separation and ramified morphology of macrophage-like cells on the ILM surface with cell translocation over time in controls. Their distribution and morphology suggest an origin of macrophage-like cells such as microglia or hyalocytes.

Interocular Symmetry of Foveal Cone Topography in Congenital Achromatopsia.

Purpose: To determine the interocular symmetry of foveal cone topography in achromatopsia (ACHM) using non-confocal split-detection adaptive optics scanning light ophthalmoscopy (AOSLO). Methods: Split-detector AOSLO images of the foveal cone mosaic were acquired from both eyes of 26 subjects (mean age 24.3 years; range 8-44 years, 14 females) with genetically confirmed CNGA3- or CNGB3-associated ACHM. Cones were identified within a manually delineated rod-free zone. Peak cone density (PCD) was determined using an 80 × 80 µm sampling window within the rod-free zone. The mean and standard deviation (SD) of inter-cell distance (ICD) were calculated to derive the coefficient of variation (CV). Cone density difference maps were generated to compare cone topography between eyes. Results: PCD (mean ± SD) was 17,530 ± 9,614 cones/mm2 and 17,638 ± 9,753 cones/mm2 for right and left eyes, respectively (p = .677, Wilcoxon test). The mean (± SD) for ICD was 9.05 ± 2.55 µm and 9.24 ± 2.55 µm for right and left eyes, respectively (p = .410, paired t-test). The mean (± SD) for CV of ICD was 0.16 ± 0.03 µm and 0.16 ± 0.04 µm for right and left eyes, respectively (p = .562, paired t-test). Cone density maps demonstrated that cone topography of the ACHM fovea is non-uniform with local variations in cone density between eyes. Conclusions: These results demonstrate the interocular symmetry of the foveal cone mosaic (both density and packing) in ACHM. As cone topography can differ between eyes of a subject, PCD does not completely describe the foveal cone mosaic in ACHM. Nonetheless, these findings are of value in longitudinal monitoring of patients during treatment trials and further suggest that both eyes of a given subject may have similar therapeutic potential and non-study eye can be used as a control.

Visual Acuity and Foveal Structure in Eyes with Fragmented Foveal Avascular Zones.

PURPOSE: To assess the frequency and impact of abnormal foveal avascular zone (FAZ) topography (i.e., a fragmented FAZ) on visual acuity and foveal anatomic features. DESIGN: Prospective, cross-sectional study from March 2018 through July 2019. PARTICIPANTS: Two-hundred fifty participants were screened from a normative OCT angiography database. Of those, 12 participants were found to have at least 1 eye with a fragmented FAZ. Eight returned for follow-up imaging, along with an additional 3 participants with ocular disease (amblyopia, autosomal recessive bestrophinopathy, premature birth) having a similar FAZ phenotype. METHODS: Follow-up OCT imaging and monocular best-corrected visual acuity (BCVA) were performed for these 11 participants. Twenty-four participants with a clearly defined FAZ were recruited for comparison. A normative database was created measuring parafoveal intercapillary area (PICA) to determine if an FAZ was fragmented. MAIN OUTCOME MEASURES: Monocular BCVA, foveal pit depth, foveal pit area, PICA, outer nuclear layer thickness, foveal inner retinal area, and peak cone density. RESULTS: The frequency of a fragmented FAZ was 4.8% of individuals (12 of 250) or 3.6% of eyes (18 of 500 eyes). A significant difference was found between the control eyes and eyes with fragmented FAZs for foveal pit depth, pit area, and total PICA (P < 0.001, P = 0.002, and P < 0.001, respectively). The presence of a fragmented FAZ did not affect visual acuity. CONCLUSIONS: The presence of a fragmented FAZ seems not to be a rare phenotype in individuals with normal vision. The presence of altered FAZ topography in patients with retinal or systemic disease could negatively impact the accuracy and sensitivity of biomarkers dependent on FAZ identification.

Foveal avascular zone morphology and parafoveal capillary perfusion in sickle cell retinopathy.

BACKGROUND/AIMS: To assess foveal avascular zone (FAZ) morphology and parafoveal capillary perfusion in patients with various stages of sickle cell retinopathy (SCR) using optical coherence tomography angiography (OCT-A). METHODS: This is a multi-institutional retrospective study of patients with various stages of SCR compared with healthy controls. Parafoveal OCT-A images obtained using a commercial spectral domain-OCT system were reviewed. Foveal-centred 3×3 mm full vascular slab OCT-As were used for image processing and data analysis. FAZ area, perimeter, and acircularity index were determined on the OCT-A image after manual delineation of the FAZ border. Quadrant-based parafoveal capillary density and per cent area deviating from normal distribution were also measured. RESULTS: Fifty-two patients with SCR (33 non-proliferative and 19 proliferative) and 20 age and race-matched healthy controls were included. One randomly selected eye per study participant was analysed. FAZ perimeter and acircularity index were significantly greater in SCR eyes when compared with the controls. While parafoveal capillary density was significantly lower, per cent area deviated from normal distribution was significantly higher in SCR eyes than that of the control. However, no statistically significant difference between the two SCR stages was observed. In quadrant-based analysis, the temporal quadrant showed greater parafoveal capillary dropout due to SCR, with the most profound effect in patients with proliferative SCR. CONCLUSIONS: Abnormal FAZ morphology and altered parafoveal capillary perfusion were found in patients with SCR. Our customised OCT-A image analysis method uniquely highlights significant quantitative alterations in perfusion density mapping in a qualitative display, with minimal obscuration of OCT-A image detail.

Assessing the Use of Incorrectly Scaled Optical Coherence Tomography Angiography Images in Peer-Reviewed Studies: A Systematic Review.

Importance: Individual differences in axial length affect the lateral magnification of in vivo retinal images and as a result can affect the accuracy of quantitative measurements made from these images. As measurements from optical coherence tomography angiography (OCTA) images are becoming increasingly used in the diagnosis and monitoring of a wide range of diseases, evaluating which studies use correctly scaled images is crucial to their interpretation. Objective: To perform a systematic literature review to assess the percentage of articles that report correcting the scale of their OCTA images for individual differences in retinal magnification. Evidence Review: A PubMed (MEDLINE) search was conducted for articles on OCTA retinal imaging published between June 1, 2015, and June 1, 2018. Initial results included 7552 articles. Initial exclusion criteria removed studies of animal models, as well as reviews, letters, replies, comments, and image-based or photographic essays. Articles not written in English and those that required purchase from non-English language websites were excluded. Articles that did not use OCTA for imaging the retina were also excluded. Remaining articles were reviewed in detail to assess whether the OCTA measurements required correct lateral scaling, and if so, whether axial length was reported or used to scale the images. We also determined the number of articles that mentioned the lack of correct lateral scaling as a limitation of the study. Findings: A total of 989 articles were included in the detailed review. Of these, 509 were determined to require correct image scaling for their analyses, but only 41 (8.0%) report measuring and using axial length to correct the lateral scale of their OCTA images. Furthermore, of the 468 articles that did not correctly scale their images, only 18 (3.8%) mentioned this as a limitation to their study. Conclusions and Relevance: These findings suggest that most peer-reviewed articles in PubMed that use quantitative OCTA measurements use incorrectly scaled images. This could call into question the conclusions of such studies and warrants consideration by OCTA manufacturers, physicians, authors, journal reviewers, and journal editors.

Interocular symmetry, intraobserver repeatability, and interobserver reliability of cone density measurements in the 13-lined ground squirrel.

BACKGROUND: The 13-lined ground squirrel (13-LGS) possesses a cone-dominant retina that is highly amenable to non-invasive high-resolution retinal imaging. The ability for longitudinal assessment of a cone-dominant photoreceptor mosaic with an adaptive optics scanning light ophthalmoscope (AOSLO) has positioned the 13-LGS to become an accessible model for vision research. Here, we examine the interocular symmetry, repeatability, and reliability of cone density measurements in the 13-LGS. METHODS: Thirteen 13-LGS (18 eyes) were imaged along the vertical meridian with a custom AOSLO. Regions of interest were selected superior and inferior to the optic nerve head, including the cone-rich visual streak. Non-confocal split-detection was used to capture images of the cone mosaic. Five masked observers each manually identified photoreceptors for 26 images three times and corrected an algorithm's cell identification outputs for all 214 images three times. Intraobserver repeatability and interobserver reliability of cone density were characterized using data collected from all five observers, while interocular symmetry was assessed in five animals using the average values of all observers. The distribution of image quality for all images in this study was assessed with open-sourced software. RESULTS: Manual identification was less repeatable than semi-automated correction for four of the five observers. Excellent repeatability was seen from all observers (ICC = 0.997-0.999), and there was good agreement between repeat cell identification corrections in all five observers (range: 9.43-25.71 cells/degree2). Reliability of cell identification was significantly different in two of the five observers, and worst in images taken from hibernating 13-LGS. Interocular symmetry of cone density was seen in the five 13-LGS assessed. Image quality was variable between blur- and pixel intensity-based metrics. CONCLUSIONS: Interocular symmetry with repeatable cone density measurements suggest that the 13-LGS is well-suited for longitudinal examination of the cone mosaic using split-detection AOSLO. Differences in reliability highlight the importance of observer training and automation of AOSLO cell detection. Cone density measurements from hibernating 13-LGS are not repeatable. Additional studies are warranted to assess other metrics of cone health to detect deviations from normal 13-LGS in future models of cone disorder in this species.

Novel Development of Parafoveal Capillary Density Deviation Mapping using an Age-Group and Eccentricity Matched Normative OCT Angiography Database.

PURPOSE: We evaluate the impact of age and signal strength index (SSI) on foveal avascular zone (FAZ) metrics and parafoveal capillary density measured using optical coherence tomography angiography (OCT-A), and propose a deviation mapping approach that accounts for age-group, SSI, eccentricity, and variation in FAZ size. METHODS: Parafoveal OCT-A with full vascular layer was obtained for 261 controls and four patients with retinal abnormalities. Parafoveal capillary densities were measured within eight consecutive 200-µm wide annuli from the FAZ border. In controls, the impacts of age and SSI on FAZ metrics and parafoveal capillary density were evaluated. Deviation maps highlighting regions with density at the lower and upper tails of the age-group and eccentricity matched distribution were generated. RESULTS: Linear regressions showed significant correlations between age, SSI, and mean parafoveal capillary density. There was a significant difference in FAZ metrics and parafoveal capillary densities with different age groups after controlling for SSI using univariate analysis. However, the effect of age on parafoveal capillary density disappeared after controlling for SSI using multivariate linear regression analysis. Our deviation mapping approach was able to identify regions with abnormal density in four patients. CONCLUSIONS: Our findings suggest that the relationship between parafoveal capillary density and age is confounded by SSI. Parafoveal capillary density is SSI- and eccentricity-dependent. An age-group and eccentricity matched normative database was used as the basis for a parafoveal capillary density deviation mapping technique, providing an intuitive way to assess the status of parafoveal capillary density in individual eyes. TRANSLATIONAL RELEVANCE: Understanding the impact of age and SSI on parafoveal capillary density is critical for providing accurate interpretation of OCT-A. We demonstrate an age-group and eccentricity matched deviation mapping technique for an intuitive assessment of retinal regions with abnormal density.

Earliest Evidence of Preclinical Diabetic Retinopathy Revealed Using Optical Coherence Tomography Angiography Perfused Capillary Density.

PURPOSE: To compare perfused capillary density (PCD) in diabetic patients and healthy controls using optical coherence tomography angiography (OCTA). METHODS: Forty controls, 36 diabetic subjects without clinical retinopathy (NoDR), 38 with nonproliferative retinopathy (NPDR), and 38 with proliferative retinopathy (PDR) were imaged using spectral-domain optical coherence tomography. A 3 × 3-mm full-thickness parafoveal OCTA scan was obtained from each participant. Following manual delineation of the foveal avascular zone (FAZ), FAZ area, perimeter, and acircularity index were determined. Seven consecutive equidistant 200-µm-wide annular segments were drawn at increasing eccentricities from the FAZ margin. Annular PCD (%) was defined as perfused capillary area divided by the corresponding annulus area after subtraction of noncapillary blood vessel areas. Nonparametric Kruskal-Wallis testing with Bonferroni correction was performed in pairwise comparisons of group PCD values. RESULTS: The NoDR group demonstrated consistently higher PCD compared to the control group in all 7 annuli, reaching statistical significance (36.6% ± 3.30% vs 33.6% ± 3.98%, P = .034) at the innermost annulus (FAZ margin to 200 µm out). The NPDR and PDR groups demonstrated progressively decreasing PCD. Differences in FAZ metrics between the NoDR and control groups did not reach statistical significance. CONCLUSIONS: Relative to healthy controls, increased PCD values in the NoDR group likely represent an autoregulatory response to increased metabolic demand, while the decrease in PCD that follows in NPDR and PDR results largely from an incremental loss of capillary segments. These findings, consistent with previous studies, demonstrate the potential of OCTA as a clinical tool for earlier objective detection of preclinical diabetic retinopathy. NOTE: Publication of this article is sponsored by the American Ophthalmological Society.

The Utility of Frame Averaging for Automated Algorithms in Analyzing Retinal Vascular Biomarkers in AngioVue OCTA.

PURPOSE: This study proposes an optical coherence tomography angiography (OCTA) frame-averaging method and investigates the effects of the number of frames acquired and averaged on metrics quantifying the foveal avascular zone (FAZ), vessel morphology, and parafoveal intercapillary area (PICA). METHODS: Ten OCTA frames were acquired for each of the 19 subjects without known retinal disease using the AngioVue OCTA system. For each subject, acquired frames were ranked by an image quality metric. A subset of frames was then registered and averaged. The effects of the number of frames acquired and averaged on FAZ segmentation and metrics of FAZ geometry, vessel morphology, and PICA were analyzed. RESULTS: Frame averaging increased the accuracy of the automatically segmented FAZ region; for example, the absolute error in FAZ area decreased from 0.026 mm2 (1 frame) to 0.005 mm2 (5 frames). Averaging multiple frames exponentially decreased the estimated number of vessel endpoints and increased the average vessel length with a 32% decrease in number of endpoints and 14% increase in average vessel length when averaging five frames compared with one. Frame averaging also improved the precision of PICA estimates. CONCLUSIONS: Averaging multiple OCTA frames using the Optovue AngioVue system reduced error in FAZ segmentation and improved the robustness of OCTA vessel morphology and perfusion metrics. The study demonstrated limited benefit in acquiring and averaging more than five frames. TRANSLATIONAL RELEVANCE: Averaging multiple OCTA frames improved the robustness of OCTA foveal biomarkers with limited benefit when averaging more than five frames.

The Henle Fiber Layer in Albinism: Comparison to Normal and Relationship to Outer Nuclear Layer Thickness and Foveal Cone Density.

Purpose: Directional optical coherence tomography (D-OCT) allows the visualization of the Henle fiber layer (HFL) in vivo. Here, we used D-OCT to characterize the HFL and outer nuclear layer (ONL) in albinism and examine the relationship between true foveal ONL and peak cone density. Methods: Horizontal D-OCT B-scans were acquired, registered, and averaged for 12 subjects with oculocutaneous albinism and 26 control subjects. Averaged images were manually segmented to extract HFL and ONL thickness. Adaptive optics scanning light ophthalmoscopy was used to acquire images of the foveal cone mosaic in 10 subjects with albinism, from which peak cone density was assessed. Results: Across the foveal region, the HFL topography was different between subjects with albinism and normal controls. In particular, foveal HFL thickness was thicker in albinism than in normal controls (P < 0.0001), whereas foveal ONL thickness was thinner in albinism than in normal controls (P < 0.0001). The total HFL and ONL thickness was not significantly different between albinism and controls (P = 0.3169). Foveal ONL thickness was positively correlated with peak cone density in subjects with albinism (r = 0.8061, P = 0.0072). Conclusions: Foveal HFL and ONL topography are significantly altered in albinism relative to normal controls. Our data suggest that increased foveal cone packing drives the formation of Henle fibers, more so than the lateral displacement of inner retinal neurons (which is reduced in albinism). The ability to quantify foveal ONL and HFL may help further stratify grading schemes used to assess foveal hypoplasia.