Abstract:We present a novel fully automated algorithm for the detection of retinal diseases via optical coherence tomography (OCT) imaging. Our algorithm utilizes multiscale histograms of oriented gradient descriptors as feature vectors of a support vector machine based classifier. The spectral domain OCT data sets used for cross-validation consisted of volumetric scans acquired from 45 subjects: 15 normal subjects, 15 patients with dry age-related macular degeneration (AMD), and 15 patients with diabetic macular edema (DME). Our classifier correctly identified 100% of cases with AMD, 100% cases with DME, and 86.67% cases of normal subjects. This algorithm is a potentially impactful tool for the remote diagnosis of ophthalmic diseases.

Endogenous fungal endophthalmitis, involving only the chorioretinal structures or extending to involve the vitreous (vitritis), is a sight-threatening infection requiring early appropriate therapy. Endophthalmitis is a relatively frequent complication of candidemia and less commonly occurs in patients who have invasive aspergillosis. Because the eye is a protected compartment, penetration of systemically administered antifungal agents is highly variable. In the posterior segment of the eye, amphotericin B (AmB) achieves very poor concentrations, but fluconazole concentrations are high. Among newer antifungal agents, voriconazole shows the most promise, because therapeutic concentrations for most Candida and Aspergillus species are achieved in the vitreous, and its antifungal activity is broad. In contrast, neither posaconazole nor the 3 echinocandins achieve adequate therapeutic concentrations in the vitreous. For sight-threatening macular involvement and vitritis, intravitreal injection of either AmB or voriconazole is helpful to achieve high local antifungal activity as quickly as possible. We review the available evidence regarding the most appropriate use of antifungal agents for endogenous fungal endophthalmitis, with the emphasis on treatment of infections due to Candida species.

Purpose To identify risk factors associated with central retinal vein occlusion (CRVO) among a diverse group of patients throughout the United States Design Longitudinal cohort study Participants All beneficiaries age ≥ 55 years who were continuously enrolled in a managed care network for at least ≥2 years and who had 2 visits to an eye care provider from 2001–2009 Methods Insurance billing codes were used to identify individuals with a newly-diagnosed CRVO. Multivariable Cox regression was performed to determine factors associated with CRVO development. Main Outcome Measure Adjusted hazard ratios (HR) with 95% confidence intervals (CI) of being diagnosed with CRVO Results Of the 494,165 enrollees who met the study inclusion criteria, 1,302 (0.26%) were diagnosed with CRVO over 5.4 (±1.8) years. After adjustment for known confounders, blacks had a 57% increased risk of CRVO compared with whites (HR = 1.57 [95% CI: 1.25–1.98]) and females had a 24% decreased risk of CRVO compared with males (HR = 0.76 [95% CI: 0.67–0.85]). A diagnosis of stroke increased the hazard of CRVO by 45% (HR = 1.45 [95% CI: 1.24–1.70]) and hypercoagulable state was associated with a 146% increased CRVO risk (HR = 2.46 [95% CI: 1.41–4.29]). Individuals with end-organ damage from hypertension (HTN) or diabetes mellitus (DM) had a 90% (HR = 1.90 [95% CI: 1.50–2.41]) and 53% (HR = 1.53 [95% CI: 1.28–1.84]) increased risk of CRVO, respectively, relative to those without these conditions. Conclusion This study confirms that HTN and vascular diseases are important risk factors for CRVO. We also identify black race as a predictor of CRVO that was not well-appreciated previously. Furthermore, we show that compared to patients without DM, individuals with end-organ damage from DM (i.e., “complicated” cases) have a heightened risk of CRVO, while those with uncomplicated DM are not at increased risk of CRVO. This finding may provide a potential explanation for the conflicting reports in the literature on the association between CRVO and DM. Information from analyses like this can be used to create a risk calculator to identify individuals at greatest risk for CRVO.