Development of LuxIA, a Cloud-Based AI Diabetic Retinopathy Screening Tool Using a Single Color Fundus Image.

Purpose: Diabetic retinopathy (DR) is the leading cause of vision impairment in working-age adults. Automated screening can increase DR detection at early stages at relatively low costs. We developed and evaluated a cloud-based screening tool that uses artificial intelligence (AI), the LuxIA algorithm, to detect DR from a single fundus image. Methods: Color fundus images that were previously graded by expert readers were collected from the Canarian Health Service (Retisalud) and used to train LuxIA, a deep-learning-based algorithm for the detection of more than mild DR. The algorithm was deployed in the Discovery cloud platform to evaluate each test set. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve were computed using a bootstrapping method to evaluate the algorithm performance and compared through different publicly available datasets. A usability test was performed to assess the integration into a clinical tool. Results: Three separate datasets, Messidor-2, APTOS, and a holdout set from Retisalud were evaluated. Mean sensitivity and specificity with 95% confidence intervals (CIs) reached for these three datasets were 0.901 (0.901-0.902) and 0.955 (0.955-0.956), 0.995 (0.995-0.995) and 0.821 (0.821-0.823), and 0.911 (0.907-0.912) and 0.880 (0.879-0.880), respectively. The usability test confirmed the successful integration of LuxIA into Discovery. Conclusions: Clinical data were used to train the deep-learning-based algorithm LuxIA to an expert-level performance. The whole process (image uploading and analysis) was integrated into the cloud-based platform Discovery, allowing more patients to have access to expert-level screening tools. Translational Relevance: Using the cloud-based LuxIA tool as part of a screening program may give diabetic patients greater access to specialist-level decisions, without the need for consultation.

Does real-time artificial intelligence-based visual pathology enhancement of three-dimensional optical coherence tomography scans optimise treatment decision in patients with nAMD? Rationale and design of the RAZORBILL study.

BACKGROUND/RATIONALE: Artificial intelligence (AI)-based clinical decision support tools, being developed across multiple fields in medicine, need to be evaluated for their impact on the treatment and outcomes of patients as well as optimisation of the clinical workflow. The RAZORBILL study will investigate the impact of advanced AI segmentation algorithms on the disease activity assessment in patients with neovascular age-related macular degeneration (nAMD) by enriching three-dimensional (3D) retinal optical coherence tomography (OCT) scans with automated fluid and layer quantification measurements. METHODS: RAZORBILL is an observational, multicentre, multinational, open-label study, comprising two phases: (a) clinical data collection (phase I): an observational study design, which enforces neither strict visit schedule nor mandated treatment regimen was chosen as an appropriate design to collect data in a real-world clinical setting to enable evaluation in phase II and (b) OCT enrichment analysis (phase II): de-identified 3D OCT scans will be evaluated for disease activity. Within this evaluation, investigators will review the scans once enriched with segmentation results (i.e., highlighted and quantified pathological fluid volumes) and once in its original (i.e., non-enriched) state. This review will be performed using an integrated crossover design, where investigators are used as their own controls allowing the analysis to account for differences in expertise and individual disease activity definitions. CONCLUSIONS: In order to apply novel AI tools to routine clinical care, their benefit as well as operational feasibility need to be carefully investigated. RAZORBILL will inform on the value of AI-based clinical decision support tools. It will clarify if these can be implemented in clinical treatment of patients with nAMD and whether it allows for optimisation of individualised treatment in routine clinical care.

Microperimetry in hydroxychloroquine macular toxicity.

Objective: The aim was to evaluate the value of microperimetry (MP) in the early detection of toxic maculopathy caused by HCQ treatment in patients with normal fundoscopy, as well as normal structural optical coherence tomography (OCT). Materials and methods: Microperimetry was performed in 13 patients under hydroxychloroquine treatment, who did not present fundoscopic or structural OCT alterations compatible with maculopathy. We used Nidek MP3s equipment (Nidek, Gamagori, Japan) with a 13-point pattern centered in fovea, in mesopic mode and in scotopic mode. Results: The mean retinal sensitivity (MRS) in the study group was 27.25 +/ - 2.80 dB (95% CI 26.09 to 28.41 dB) while in the group of healthy volunteers 29.34 +/ - 2.18 dB (95% CI 28.67 to 30.1 dB). In scotopic mode, the mean sensitivity was 13.38 +/ - 1.43 dB (95% CI 12.79 to 13.97 dB) for HCQ users and 14.40 +/ - 2.1 dB (95% CI 13.76 to 15.04 dB) in the non-user group. Central retinal sensitivity (CRS) was also lower in patients using HCQ 26.52 +/ -4.0 dB (95% CI 24.8 to 28.15 dB) vs. 29.06 +/ - 2.5 dB (95% CI 28.33 to 29.87 dB) in the control group in mesopic mode. The trend was repeated in scotopic CRS (10.85 +/ -1.84 dB vs. 12.16 +/ - 2.61 dB respectively). Discussion: Our results showed that MP, especially in its mesopic mode, is a useful method to detect retinal toxicity caused by HCQ consumption in patients without funduscopic alteration and with normal macular OCT. Conclusions: In mesopic mode, MRS was significantly lower in patients with long-term hydroxychloroquine treatment compared to those who did not use it, even in cases in which no fundoscopic or structural OCT alteration was detected. Abbreviations: HCQ = hydroxychloroquine, MP = microperimetry, OCT = optical coherence tomography, BCVA = best corrected visual acuity, CRS = central retinal sensitivity, RS = retinal sensitivity, GEE = generalized estimating equations, MRS = mean retinal sensitivity, MfERG = multifocal electroretinogram, AMD = age-related macular degeneration.

Structure-Function Correlation Using OCT Angiography And Microperimetry In Diabetic Retinopathy.

PURPOSE: To evaluate macular vasculature in diabetic retinopathy (DR) with optical coherence tomography angiography (OCTA) and to correlate vessel density (VD) with retinal sensitivity (RS) as a way to assess structural and functional findings in DR. DESIGN: Prospective observational cross-sectional study. METHODS: Diabetic patients with DR but no clinically significant diabetic macular edema (DME) and healthy subjects were included in this study. All of them underwent comprehensive ophthalmic examination, best corrected visual acuity (BCVA), OCTA with RS-3000 Advance AngioScan (Nidek, Gamagori, Japan) and microperimetry with MP-3 (Nidek, Gamagori, Japan). Retinal vascular density measured by OCT angiography in 9 areas was correlated with RS in the same 9 areas by Spearman correlation. RESULTS: In this study, 50 subjects were enrolled: 25 eyes of diabetic patients with DR and 25 eyes of non-diabetic subjects. Diabetic patients mean age was 51.88±13.62 years; non-diabetic subjects were 43.48±13.42 years. The BCVA was 20/25 in the diabetic group and 20/20 in the non-diabetic group. Mean RS was decreased in the DR group (27.68±2.71 dB) compared to the non-diabetic group (31.68±1.46 dB) (p<0.05) and in the 9 studied areas (p<0.05). Mean VD was decreased in the DR group compared to non-diabetics (p<0.05) and in 7 of the 9 areas (except temporal superior and inferior squares) (p<0.05). Correlations by areas between VD and RS were assessed, we found moderate correlation in the area temporal to the fovea (r=0.501, p=0.01) in the DR group but not in the non-diabetic subjects. There were no other statistical significant correlations with this pattern. CONCLUSION: DR without DME results in a retinal VD decreased that might be the cause of a reduction of RS in one of the studied areas. Microangiopathic changes are correlated with microperimetry sensitivity drop in the temporal to the fovea grid but not in the other studied grids.

Management of diabetic macular edema patients in clinical practice in Spain.

PURPOSE: Diabetic macular edema is the main cause of blindness in diabetic patients. Vascular endothelial growth factor is involved in diabetic macular edema pathogenesis. Vascular endothelial growth factor inhibitors are an important option in diabetic macular edema therapy. This survey investigates actual clinical practice in diabetic macular edema in Spain. METHODS: An expert advisory panel of 17 Spanish ophthalmologists developed a 30-item anonymous questionnaire about diagnosis, treatment, and follow-up in diabetic macular edema. A total of 137 ophthalmologists from 10 Spanish regions completed the questionnaire online. RESULTS: Almost all of the respondents (99.3%) record the measured visual acuity and perform biomicroscopic anterior (94.9%) and posterior (91.2%) segment examinations. Similarly, 100% of responding ophthalmologists always/almost always or frequently perform optical coherence tomography. Most respondents (65%) always/almost always or frequently perform a retinography. More than 50% rarely perform fluorescein angiography. Nearly, all (96.4%) of the specialists responded that, in center-involved diabetic macular edema, the first treatment is an anti-vascular endothelial growth factor drug. For corticosteroids, the first choice of most respondents (91.2%) was the dexamethasone implant. In the follow-up, almost all (96.4%) specialists record the measured visual acuity and most also perform biomicroscopic anterior (82.5%) and posterior (83.2%) segment examination. CONCLUSION: This survey shows the actual clinical practice in diabetic macular edema in Spain, finding that anti-vascular endothelial growth factor therapy is frequently used, and that diagnosis, treatments, and follow-up examinations used by specialists are homogeneous and according to diabetic macular edema guidelines.

New visual field indices of disharmony for early diagnosis of glaucoma, alone or associated with conventional parameters.

PURPOSE: To evaluate the specificity of new perimetric indices based on harmony, alone and in combination with structural data, for glaucoma detection. METHODS: In this prospective observational cross-sectional study, one eye of 105 healthy subjects and 97 early and suspect glaucomas were sequentially included and examined with Cirrus optical coherence tomography, twice with OCULUS Smartfield perimeter (SPARK strategy) and twice with Humphrey Analyzer (24-2 SITA-Fast) at the Ophthalmology Department from the University Hospital La Candelaria. Disharmony in the visual field was evaluated including vertical threshold symmetry, threshold rank), and homogeneity (threshold standard deviation from its maximum) using the patient himself/herself as a reference. We also evaluated disharmony in combination with the mean deviation and the pattern standard deviation in a single index (mismatch) and various combinations of morphological and functional indices. Combinations used a new score based on values above certain critical cut-off levels of each index. RESULTS: For 95% specificity, the highest sensitivities were as follows: vertical cup/disc ratio: 28.9%; SPARK threshold rank: 29.9%; and SITA-Fast threshold standard deviation: 28.9%. For the combined indices and 100% specificity, they were 5 SPARK indices mismatch: 10.3%; 5 SITA-Fast indices mismatch: 11.3%; 8 optical coherence tomography indices: 21.9%; 13 SPARK and optical coherence tomography indices: 27.8%; and 13 SITA-Fast and optical coherence tomography indices: 32.0%. CONCLUSION: Disharmony combined with normative value-based indices and/or optical coherence tomography indices is useful for very specific early diagnosis of glaucoma.

Intracameral cefuroxime and moxifloxacin used as endophthalmitis prophylaxis after cataract surgery: systematic review of effectiveness and cost-effectiveness.

Postoperative endophthalmitis is one of the most serious potential complications of ocular lens surgery. Its incidence can be reduced by means of antibiotic prophylaxis. Although the prophylactic use of intracameral cefuroxime has been extended, other drugs, such as moxifloxacin, have arisen as alternatives. We performed a systematic literature review on the effectiveness and efficiency of intracameral cefuroxime and moxifloxacin for the prophylaxis of postoperative endophthalmitis after cataract surgery. Several bibliographic databases were searched up to October 2010 and were updated up to January 2013. Outcomes were the onset of endophthalmitis after surgery and the cost-effectiveness ratio of using both antibiotic prophylaxis alternatives. The following were included: a clinical trial reported in two papers, six observational studies, and an economic evaluation. All studies assessed cefuroxime compared with another antibiotic prophylaxis or no prophylaxis. The only randomized controlled trial performed by the European Society of Cataract and Refractive Surgery found that intracameral cefuroxime is significantly more effective than not using prophylaxis or the use of a topical antibiotic. The observational studies support these results. The economic evaluation compared different prophylaxis regimens and concluded that intracameral cefuroxime showed the best cost-effectiveness ratio. Both the observational studies and the economic evaluation have methodological limits that reduce their validity. This review confirmed that cefuroxime can prevent endophthalmitis after cataract surgery. Further randomized controlled trials, with large sample sizes, are required to compare different antibiotic prophylaxis regimens.

Pigment dispersion syndrome associated with intraocular lens implantation: a new surgical technique.

AIMS: We report the case of a myopic patient who, after intraocular lens transplant in the posterior chamber, suffered elevated intraocular pressure due to pigment dispersion, with recurrent episodes of blurred vision. The patient was treated with a new surgical technique that can avoid potential iridolenticular contact. METHODS: Complete ophthalmologic examination and optical coherence tomography (OCT) of the anterior segment were performed. RESULTS: Contact between the pigmentary epithelium and the iris with an intraocular lens was revealed by utrasound biomicroscopy and OCT. In this case, Nd:YAG laser iridotomy and laser iridoplasty were not effective for iridolenticular separation and control of the pigment dispersion. We propose a new technique: stitches on the surface of the iris to obtain good iridolenticular separation and good intraocular pressure control. CONCLUSION: Stitches on the iris surface should be considered as optional therapy in pigmentary glaucoma secondary to intraocular lens implantation. This surgical technique can avoid potential iridolenticular contacts more definitively.