Faculty Bibliography

BACKGROUND AND OBJECTIVE/OBJECTIVE:To use quantitative fundus autofluorescence (qAF) to analyze different stages of non-neovascular age-related macular degeneration (AMD). PATIENTS AND METHODS/METHODS:In this cohort study, 38 pseudophakic patients and 36 age-matched controls participated. We performed near-infrared, spectral-domain optical coherence tomography and qAF imaging on 31 pseudophakic eyes and controls of participants older than 60 years with non-neovascular AMD phenotypes using the Spectralis HRA + OCT (Heidelberg Engineering, Heidelberg, Germany). RESULTS:The patients included in this study had a mean age of 83.9 years, and 35.7% patients were men. Mean qAF was higher in control participants than in all patients with AMD (P < .001). According to non-neovascular AMD phenotype, mean qAF levels were significantly lower in eyes with subretinal drusenoid deposits than in control eyes (P < .05). The lowest mean qAF was in patients with geographic atrophy. CONCLUSION/CONCLUSIONS:Quantitative fundus autofluorescence of non-neovascular AMD decreases from normal to early to late AMD, suggesting that loss of lipofuscin fluorophores, not increase, signifies AMD progression. [Ophthalmic Surg Lasers Imaging Retina. 2018;49:S34-S42.].

To characterize fluorophore signals from drusen and retinal pigment epithelium (RPE) and their changes in age related macular degeneration (AMD), the authors describe advances in ex vivo hyperspectral autofluorescence (AF) imaging of human eye tissue. Ten RPE flatmounts from eyes with AMD and 10 from eyes without AMD underwent 40× hyperspectral AF microscopic imaging. The number of excitation wavelengths tested was initially two (436 nm and 480 nm), then increased to three (436 nm, 480 nm, and 505 nm). Emission spectra were collected at 10 nm intervals from 420 nm to 720 nm. Non-negative matrix factorization (NMF) algorithms decomposed the hyperspectral images into individual emission spectra and their spatial abundances. These include three distinguishable spectra for RPE fluorophores (S1, S2, and S3) in both AMD and non-AMD eyes, a spectrum for drusen (SDr) only in AMD eyes, and a Bruch's membrane spectrum that was detectable in normal eyes. Simultaneous analysis of datacubes excited atthree excitation wavelengths revealed more detailed spatial localization of the RPE spectra and SDr within drusen than exciting only at two wavelengths. Within AMD and non-AMD groups, two different NMF initialization methods were tested on each group and converged to qualitatively similar spectra. In AMD, the peaks of the SDr at ~510 nm (436 nm excitation) were particularly consistent. Between AMD and non-AMD groups, corresponding spectra in common, S1, S2, and S3, also had similar peak locations and shapes, but with some differences and further characterization warranted.

In this paper, we provide a new framework on deep learning based automated screening method for finding individuals at risk of developing Age-related Macular Degeneration (AMD). We studied the appropriateness of using the transfer learning to screen AMD by using color fundus images. We make use of the Age-Related Eye Disease Study (AREDS) dataset with nearly 150,000 images, which also provided qualitative grading information by expert graders and ophthalmologists. We use ensemble learning technique with two deep neural networks, namely, Inception-ResNet-V2 and Xception with a custom fine-tuning approach. For our study, we have identified two experiments that are most useful in the screening of AMD. First, we have categorized the images into two classes based on the clinical significance: None or early AMD and Intermediate or Advanced AMD. Second, we have categorized the images into four classes: No AMD, early AMD, Intermediate AMD and Advanced AMD. On AREDS dataset, we have achieved an accuracy of over 95.3% for two-class experiment with our ensemble method. With accuracies ranging from 86% (for four-class) to 95.3% (for two-class), we have demonstrated that the training of a deep neural network with the transfer of learned features with a sufficient number of images fares very well and is comparable to human grading.

Purpose: Bestrophinopathies in humans and canine Best (cBest) disease both result from mutations of the BEST1 gene, and share certain features including presentation of retinal lesions. Hyperspectral autofluorescence (AF) imaging in humans ex vivo reveals characteristic signatures in normal eyes and those with age-related macular degeneration (AMD). Similar analysis of cBest fluorophore spectra and distribution in comparison could elucidate understanding of the physiology of both species.
Method(s): Two cBest and two normal canine specimens yielded four tissue cross sections that were captured using 20X FIeld hyperspectral AF imaging (Fig 1a). Excitation wavelengths of 436 480, and 505 nm were used with emissions collected from 420 nm to 720 nm, and the resulting multi-excitation hyperspectral data was unmixed using a nonnegative matrix factorization into fluorophore emission spectra and spatial distributions.
Result(s): Two of the RPE fluorophore spectral proFIles recovered in both normal dogs and cBest (Fig. 1b, spectra C2 and C5) matched those previously described in normal human eyes as S1 (or S1B) and S2 (Ben Ami et al., PMID 27226929). In some samples, only S1B was recovered. These fluorophores localized to both RPE and cBest lesions (Figs. 1c, 1d), instead of just within the RPE as in humans. There was another spectra that corresponded to both weak non-RPE autofluorescence and background noise (C1).
Conclusion(s): The presence of human RPE spectra in canine retina, in both normal and cBest samples, offers insight into the physiology of both species, and may apply to human Best disease. The physical distribution in cBest to both RPE and vitelliform lesions warrants further inquiry. (Figure presented)

Purpose: Mutations of bestrophin 1 (BEST1) cause retinal disorders such as bestrophinopathies in human and canine BEST1 disease (cBest1) in dogs. The spectrum of lipofuscin (LF) in cBest1 and normal dogs has been shown to be the same in PMID:28111324. The further characterization of fluorophores in the RPE and in the vitelliform material of cBest1 may extend our understanding of pathology of cBest1 and more information on the human disease as well.
Method(s): We prepared cross section slides from a canine retina with cBest1. We acquired images using a Nuance hyperspectral camera connected to a Zeiss microscope with 20X objective and 10nm emission intervals from 420 nm to 720 nm at excitations 436 nm, 480 nm and 505 nm. We applied non-negative matrix factorization for image processing to extract 5 abundances and their corresponding spectra.
Result(s): Fig. 1 is a combined RGB AF image of a cBest1 retina. Recovered spectra C1 and C3 had peaks at 540 nm and 570 nm, respectively, with a notch at 620 nm in C3, at excitation 436 nm (Fig. 2), strikingly similar to normal human spectra S1 and S2 (PMID:27226929). The abundance images 1 and 3 were localized similarly both to the RPE and granules in vitelliform space. The granules, which appear more orange in total AF, show a greater preponderance of the longer wavelength C3 in the vitelliform debris. The spectrum C2 peaking at 510 nm localized to all tissue including collagen in the choroid, and is a background artifact. The short wavelength spectrum C5 (purple line, peak 430 nm) localized to a single vitelliform granule, abundance 5 in Fig. 2 and is consistent with fluorophores from the neural retina, such as NADP and FAD, not LF. The low intensity spectrum C4 appears to be noise.
Conclusion(s): We report 2 individual AF spectra appearing in both LF and vitelliform material in the cBest1 retina that are quite similar to human RPE spectra. The longer wavelength source predominates in the vitelliform material. These findings will enhance our understanding of the mechanism of cBest1 as we move closer to treatment of bestrophinopathies. Further research with additional tissues is warranted. (Figure presented)

Mounting evidence supports a genetic-vascular-inflammatory etiology of schizophrenia. The retina provides an indirect assessment of inflammation and degeneration in the brain. In particular, the use of spectral domain optical coherence tomography (SD-OCT) has emerged as a powerful tool for examining single retinal nerve cell layers and the choroid, the vascular layer supplying the outer retina. In this study, choroidal and macular thicknesses were measured in six patients with psychosis with either schizophrenia or bipolar disorder, and in 18 age- and sex-matched healthy controls. Mean choroidal thickness was reduced in psychosis, though not significantly so. There was a statistically significant decrease in macular thickness in psychosis patients predominantly affecting the inner layers of the macula. Significant macular thinning may signal vascular, inflammatory, or degenerative processes that may also be occurring in the brain. This is one of the first studies to examine choroidal thickness in psychosis. Further studies are needed to determine whether the retinal changes in psychosis are correlated with microvascular dysfunction, neuroinflammation, and neurodegeneration.

OBJECTIVE: This retrospective cohort study utilized 3 imaging modalities to analyze quantitatively reticular pseudodrusen (RPD) area changes in eyes that progressed from early to late age-related macular degeneration (AMD). METHODS: Subjects with AMD, unilateral choroidal neovascularization (CNV), and early AMD with RPD in the fellow eye (the study eye) were included. The study eyes underwent indocyanine green angiography (ICGA), near-infrared reflectance (NIR-R), and short-wavelength autofluorescence (AF) imaging of the macula at baseline and at follow-up. Study eyes were analyzed for RPD and for the development of late AMD-CNV and/or geographic atrophy (GA). RPD area was measured at baseline and at follow-up as a percentage of the 30-degree field. RESULTS: During the study period (mean follow-up time 23.5 +/- 5.0 months), 12/31 study eyes developed CNV and 4/31 developed GA. In the eyes that developed CNV, there was a statistically significant decrease in mean RPD area over the follow-up period as seen on AF (P < 0.01) and NIR-R (P = 0.01), and the decrease in mean RPD area approached statistical significance on ICGA (P = 0.08). CONCLUSION: Using 3 en face imaging techniques, we demonstrate that RPD undergo dynamic spatiotemporal changes in eyes that progress from early AMD to CNV, namely a decrease in the area of lesions detected.

Introduction Patient portals or personal health records allow patients to access their health information and communicate with their physician"™s office outside of their clinic visit. As such, their use has been observed to increase patient engagement and decrease administrative costs. Despite these advantages, patient adoption and successful use of patient portals remains low. Here we assess the feasibility and efficacy of an iBooks-based tutorial in increasing comfort and compliance with personal health record use. Methods We created and published a 5-min iBooks-based tutorial describing our institution"™s patient portal features. We administered the tutorial, along with presurvey and postsurvey, to ophthalmology clinic patients. Results Of 70 participants, 50% had already registered for our institution"™s patient portal. Registered patients had viewed labs (74%, n=26) and messaged providers (57%, n=20) but rarely used other features such as scheduling appointments (17%, n=6) or requesting refills (26%, n=9). After the tutorial, comfort levels in knowledge on how to use portal features increased by 20%-80%, depending on portal feature and registration status. Main barriers to portal usage were preference for telephonebased communication (26%, n=18) and knowledge of portal existence (21%, n=15). The majority (86%, n=60) agreed the tutorial would increase their utilisation of the patient portal. Conclusion Tutorials increase knowledge and awareness of patient portal features, allowing these features to be fully used. An iBooks-based approach allows patients to successfully access and use tutorial content outside of the clinic.