Moderate Diabetic Retinopathy Research Articles

Evaluation of Automatically Quantified Foveal Avascular Zone Metrics for Diagnosis of Diabetic Retinopathy Using Optical Coherence Tomography Angiography.

PurposeTo describe an automated algorithm to quantify the foveal avascular zone (FAZ), using optical coherence tomography angiography (OCTA), and to compare its performance for diagnosis of diabetic retinopathy (DR) and association with best-corrected visual acuity (BCVA) to that of extrafoveal avascular area (EAA).MethodsWe obtained 3 × 3-mm macular OCTA scans in diabetic patients with various levels of DR and healthy controls. An algorithm based on a generalized gradient vector flow (GGVF) snake model detected the FAZ, and metrics assessing FAZ size and irregularity were calculated. We compared the automated FAZ segmentation to manual delineation and tested the within-visit repeatability of FAZ metrics. The correlations of two conventional FAZ metrics, two novel FAZ metrics, and EAA with DR severity and BCVA, as determined by Early Treatment Diabetic Retinopathy Study (ETDRS) charts, were assessed.ResultsSixty-six eyes from 66 diabetic patients and 19 control eyes from 19 healthy participants were included. The agreement between manual and automated FAZ delineation had a Jaccard index > 0.82, and the repeatability of automated FAZ detection was excellent in eyes at all levels of DR severity. FAZ metrics that incorporated both FAZ size and shape irregularity had the strongest correlation with clinical DR grade and BCVA. Of all the tested OCTA metrics, EAA had the greatest sensitivity in differentiating diabetic eyes without clinical evidence of retinopathy, mild to moderate nonproliferative DR (NPDR), and severe NPDR to proliferative DR from healthy controls.ConclusionsThe GGVF snake algorithm tested in this study can accurately and reliably detect the FAZ, using OCTA data at all DR severity grades, and may be used to obtain clinically useful information from OCTA data regarding macular ischemia in patients with diabetes. While FAZ metrics can provide clinically useful information regarding macular ischemia, and possibly visual acuity potential, EAA measurements may be a better biomarker for DR.

Grader Variability and the Importance of Reference Standards for Evaluating Machine Learning Models for Diabetic Retinopathy

Use adjudication to quantify errors in diabetic retinopathy (DR) grading based on individual graders and majority decision, and to train an improved automated algorithm for DR grading. Retrospective analysis. Retinal fundus images from DR screening programs. Images were each graded by the algorithm, U.S. board-certified ophthalmologists, and retinal specialists. The adjudicated consensus of the retinal specialists served as the reference standard. For agreement between different graders as well as between the graders and the algorithm, we measured the (quadratic-weighted) kappa score. To compare the performance of different forms of manual grading and the algorithm for various DR severity cutoffs (e.g., mild or worse DR, moderate or worse DR), we measured area under the curve (AUC), sensitivity, and specificity. Of the 193 discrepancies between adjudication by retinal specialists and majority decision of ophthalmologists, the most common were missing microaneurysm (MAs) (36%), artifacts (20%), and misclassified hemorrhages (16%). Relative to the reference standard, the kappa for individual retinal specialists, ophthalmologists, and algorithm ranged from 0.82 to 0.91, 0.80 to 0.84, and 0.84, respectively. For moderate or worse DR, the majority decision of ophthalmologists had a sensitivity of 0.838 and specificity of 0.981. The algorithm had a sensitivity of 0.971, specificity of 0.923, and AUC of 0.986. For mild or worse DR, the algorithm had a sensitivity of 0.970, specificity of 0.917, and AUC of 0.986. By using a small number of adjudicated consensus grades as a tuning dataset and higher-resolution images as input, the algorithm improved in AUC from 0.934 to 0.986 for moderate or worse DR. Adjudication reduces the errors in DR grading. A small set of adjudicated DR grades allows substantial improvements in algorithm performance. The resulting algorithm's performance was on par with that of individual U.S. Board-Certified ophthalmologists and retinal specialists.

Automated diabetic retinopathy detection in smartphone-based fundus photography using artificial intelligence

ObjectivesTo assess the role of artificial intelligence (AI)-based automated software for detection of diabetic retinopathy (DR) and sight-threatening DR (STDR) by fundus photography taken using a smartphone-based device and validate it against ophthalmologist’s grading.MethodsThree hundred and one patients with type 2 diabetes underwent retinal photography with Remidio ‘Fundus on phone’ (FOP), a smartphone-based device, at a tertiary care diabetes centre in India. Grading of DR was performed by the ophthalmologists using International Clinical DR (ICDR) classification scale. STDR was defined by the presence of severe non-proliferative DR, proliferative DR or diabetic macular oedema (DME). The retinal photographs were graded using a validated AI DR screening software (EyeArtTM) designed to identify DR, referable DR (moderate non-proliferative DR or worse and/or DME) or STDR. The sensitivity and specificity of automated grading were assessed and validated against the ophthalmologists’ grading.ResultsRetinal images of 296 patients were graded. DR was detected by the ophthalmologists in 191 (64.5%) and by the AI software in 203 (68.6%) patients while STDR was detected in 112 (37.8%) and 146 (49.3%) patients, respectively. The AI software showed 95.8% (95% CI 92.9–98.7) sensitivity and 80.2% (95% CI 72.6–87.8) specificity for detecting any DR and 99.1% (95% CI 95.1–99.9) sensitivity and 80.4% (95% CI 73.9–85.9) specificity in detecting STDR with a kappa agreement of k = 0.78 (p < 0.001) and k = 0.75 (p < 0.001), respectively.ConclusionsAutomated AI analysis of FOP smartphone retinal imaging has very high sensitivity for detecting DR and STDR and thus can be an initial tool for mass retinal screening in people with diabetes.

Prevalence of diabetic retinopathy among children with type 1 diabetes mellitus treated by insulin

Purpose Among children and young adults, type 1 diabetes mellitus (DM) is the commonest type to occur as a result of defects in insulin secretion and/or insulin action. It is usually treated by insulin or insulin analogs. This work was done to study the prevalence of diabetic retinopathy (DR) among diabetic children with type 1 DM treated with insulin. Patients and methods This is a cross-sectional study that was carried out in Damnhour Teaching Hospital. It included 115 children with type 1 DM from 6 to 14 years of age. All patients were subjected to full fundus examination and were then classified according to its results into patients with DR and patients without DR. Risk factors associated with retinopathy were assessed in those patients. Results The mean age of the patients was 9.70±1.94 years and 55.7% were males. DR was found in 12 (10.4%) patients, eight of whom had retinopathy in one eye and four had retinopathy in both eyes. Mild nonproliferative DR was found in 11/230 (4.8%) eyes and moderate nonproliferative DR was found in 5/230 (2.2%) eyes. A significant positive correlation was found between the grade of retinopathy and the duration of diabetes (r=0.37; P Conclusion The prevalence of DR was 10.4% among all studied patients and 7.0% among all studied eyes. The grade of retinopathy was directly related to the duration of DM, fasting blood glucose, and glycosylated hemoglobin levels.

Lack of Longitudinal Association Between Thiazolidinediones and Incidence and Progression of Diabetic Eye Disease: The ACCORD Eye Study

To report the longitudinal association between use of thiazolidinediones (TZDs), visual acuity (VA) change, and diabetic eye disease incidence and progression. Cohort study ancillary to a randomized clinical trial. We analyzed baseline and 4-year follow-up data of 2856 ACCORD trial participants with no history of proliferative diabetic retinopathy. Based on stereoscopic fundus photographs, we evaluated diabetic macular edema (DME) progression and DR progression. We also evaluated 10- and 15-letter change on the ETDRS visual acuity chart. Main outcome measures were incidence or progression of DME or DR and change in visual acuity. TZD use was not associated with DME incidence in either the analysis of any use (adjusted odds ratio [aOR] [95% CI]: 1.22 [0.72-2.05]) or duration of use (aOR: 1.02 [0.99-1.04]). Diabetic retinopathy (DR) incidence/progression was more common in patients with no or mild DR at baseline who were ever treated with TZDs (aOR: 1.68 [1.11-2.55]), but this association disappeared when adjusting for the time on TZD (aOR: 1.02 [1.00-1.04]). DR progression among those with moderate or worse DR at baseline was no different between TZD users and non-users. TZD usage had no effect on the ultimate visual acuity outcome. In this longitudinal study of patients with type 2 diabetes, we found no association between TZD use and visual acuity outcomes or DME progression, and no consistent evidence of increased DR progression in patients ever treated with TZDs vs those never treated with TZDs.

Association Between the Severity of Diabetic Retinopathy and Falls in an Asian Population With Diabetes

The presence and severity of diabetic retinopathy (DR) may contribute to the risk of falling in persons with diabetes, but evidence is currently equivocal. To investigate the associations of diabetes and DR severity with the likelihood of falls in a multiethnic Asian population. Cross-sectional post hoc analysis of the Singapore Epidemiology of Eye Diseases study, a population-based study of participants from 3 ethnic groups (3280 Malay, 3400 Indian, and 3353 Chinese individuals) conducted from 2004 to 2011. Of these participants, 552 had data missing on diabetes, falls history, or other covariates or had ungradable fundus photographs and were excluded, leaving 9481 participants. These 9481 underwent a standardized clinical examination and responded to an interviewer-administered questionnaire that collected clinical and sociodemographic information. Multivariable logistic regression models adjusted for confounding fall risk factors assessed the associations of falls with diabetes, DR, and DR severity. A trend analysis was conducted in participants with diabetes to assess if risk of falling was associated with DR severity. Data were analyzed from January 1 through April 30, 2017. Diabetes was defined as a random glucose level of at least 200 mg/dL, hemoglobin A1c concentration of at least 6.5% of total hemoglobin, self-reported use of diabetic medication, or history of physician-diagnosed diabetes. Severity of DR was graded as none, minimal, mild, moderate, and vision threatening (VT). A self-reported fall occurring in the previous 12 months, when the participant fell and landed on the ground. Of the 9481 participants with a mean (SD) age of 58.7 (10.3) years (4781 women [50.4%]), 6612 (69.7%) had no diabetes and 2869 (30.3%) had diabetes, of whom 857 (29.9%) had DR in at least 1 eye. A history of falls was reported in 872 (13.2%) without diabetes, 328 (16.3%) with no DR, 44 (14.2%) with minimal DR, 54 (26.2%) with mild DR, 34 (27.2%) with moderate DR, and 43 (19.9%) with VTDR (P for trend < .001). In multivariable models, those with DR were more likely to have fallen (odds ratio [OR], 1.31; 95% CI, 1.07-1.60; P = .008) compared with those with no diabetes; no associations were found for participants without DR compared with those with no diabetes. In addition, compared with participants with diabetes but without DR, those with mild (OR, 1.81; 95% CI, 1.23-2.67; P = .003) and moderate (OR, 1.89; 95% CI, 1.16-3.07; P = .01) nonproliferative DR were more likely to have fallen. Having VTDR was not independently associated with a higher likelihood of falling. The presence of mild to moderate nonproliferative DR was independently associated with an increased likelihood of falling in persons with diabetes compared with persons with diabetes but without DR. Management strategies for diabetes should incorporate fall education and prevention information, particularly in patients with early-stage DR. Longitudinal studies exploring the association between mild to moderate nonproliferative DR and falling will be required to confirm these findings.

Associations of Physical Activity and Sedentary Behaviour with Vision-Threatening Diabetic Retinopathy in Indonesian Population with Type 2 Diabetes Mellitus: Jogjakarta Eye Diabetic Study in the Community (JOGED.COM)

ABSTRACTPurpose: To investigate the association of physical activity (PA) and sedentary behaviour with diabetic retinopathy (DR) in type 2 diabetes adults.Methods: A community-based cross-sectional study of adults with type 2 diabetes. DR was assessed from disc- and macula-centred fundus photograph, and categorised into mild, moderate non-proliferative DR (NPDR), and Vision-Threatening DR (VTDR). Worse eye was used to determine the person’s DR severity. PA was assessed using WHO Global Physical Activity Questionnaire (GPAQ). All of the PA components, including daily sitting time, walking distance, sleeping duration, and exercise were converted based on the physical activity ratio and then calculated to obtain the mean physical activity level (PAL). The mean PAL was then classified into sedentary, moderate, and vigorous lifestyle.Results: There were 1116 participants, 43.1% with any DR and 26.3% VTDR. The prevalence of DR/VTDR in sedentary, moderate, and vigorous physical activity group was 41.9%/22.7%, 39.2%/23.9%, and 32.7%/20.4% respectively. A longer duration of daily sedentary lifestyle (>3 hours) was significantly associated with presence of DR (OR 1.66; 95% CI 1.17–2.35; p < 0.01) and VTDR (OR 1.74; 1.16–2.62; p < 0.01). These associations remained significant after adjusting for age, gender, duration of diabetes, systolic blood pressure and physical activity. There was no significant association between sleeping duration, walking distance, and vigorous activity and DR.Conclusions: In this study, prolonged daily sedentary activity was strongly associated with VTDR. This finding suggests that public health campaign for reducing daily sedentary activities in individuals with type 2 diabetes may offer benefit to reduce the burden of DR and VTDR.

Diabetic retinopathy screening using deep neural network.

There is a burgeoning interest in the use of deep neural network in diabetic retinal screening. To determine whether a deep neural network could satisfactorily detect diabetic retinopathy that requires referral to an ophthalmologist from a local diabetic retinal screening programme and an international database. Retrospective audit. Diabetic retinal photos from Otago database photographed during October 2016 (485 photos), and 1200 photos from Messidor international database. Receiver operating characteristic curve to illustrate the ability of a deep neural network to identify referable diabetic retinopathy (moderate or worse diabetic retinopathy or exudates within one disc diameter of the fovea). Area under the receiver operating characteristic curve, sensitivity and specificity. For detecting referable diabetic retinopathy, the deep neural network had an area under receiver operating characteristic curve of 0.901 (95% confidence interval 0.807-0.995), with 84.6% sensitivity and 79.7% specificity for Otago and 0.980 (95% confidence interval 0.973-0.986), with 96.0% sensitivity and 90.0% specificity for Messidor. This study has shown that a deep neural network can detect referable diabetic retinopathy with sensitivities and specificities close to or better than 80% from both an international and a domestic (New Zealand) database. We believe that deep neural networks can be integrated into community screening once they can successfully detect both diabetic retinopathy and diabetic macular oedema.

An Automated System for the Detection and Classification of Retinal Changes Due to Red Lesions in Longitudinal Fundus Images.

People with diabetes mellitus need annual screening to check for the development of diabetic retinopathy (DR). Tracking small retinal changes due to early diabetic retinopathy lesions in longitudinal fundus image sets is challenging due to intra- and intervisit variability in illumination and image quality, the required high registration accuracy, and the subtle appearance of retinal lesions compared to other retinal features. This paper presents a robust and flexible approach for automated detection of longitudinal retinal changes due to small red lesions by exploiting normalized fundus images that significantly reduce illumination variations and improve the contrast of small retinal features. To detect spatio-temporal retinal changes, the absolute difference between the extremes of the multiscale blobness responses of fundus images from two time points is proposed as a simple and effective blobness measure. DR related changes are then identified based on several intensity and shape features by a support vector machine classifier. The proposed approach was evaluated in the context of a regular diabetic retinopathy screening program involving subjects ranging from healthy (no retinal lesion) to moderate (with clinically relevant retinal lesions) DR levels. Evaluation shows that the system is able to detect retinal changes due to small red lesions with a sensitivity of at an average false positive rate of 1 and 2.5 lesions per eye on small and large fields-of-view of the retina, respectively.

Multifocal electroretinography changes at the 1-year follow-up in a cohort of diabetic macular edema patients treated with ranibizumab

PurposeTo determine the changes in the multifocal electroretinogram (mfERG) at 1 year in a clinical series of diabetic macular edema (DME) patients treated with ranibizumab (RNBZ) using a pro re nata protocol.MethodsWe analyzed a clinical series of 35 eyes of 35 patients with DME at baseline and after treating them with RNBZ over 1 year, in order to determine the change in the macular function, which was assessed by means of the response density and the implicit time of the first-order kernel (FOK) P1 wave of the mfERG at the foveola (R1), fovea (R2) and parafovea (R3). These electrophysiological parameters were studied taking into account different independent variables, such as DME type, degree of diabetic retinopathy (DR), level of preservation of both the ellipsoid zone (IS/OS) and the external limiting membrane (ELM) and changes in central retinal thickness (CRT) and total macular volume (TMV). We also studied the relationship between the response density and the best-corrected visual acuity (BCVA).ResultsEyes with cystic and spongiform DME showed better response density with respect to the serous type (p < 0.001) at baseline. Similarly, eyes with high IS/OS and ELM preservation rates showed higher initial response density compared to the others (p < 0.001). Eyes with moderate DR had better response density compared to those with severe and proliferative DR (p = 0.001). At the beginning of the study, those eyes with proliferative and severe DR showed longer implicit times with respect to those with moderate DR (p = 0.04). The response density significantly increased in eyes that anatomically restored the IS/OS and the ELM after being treated with RNBZ (both p < 0.001). Similarly, eyes with spongiform DME further improved the response density with respect to those with cystic and serous DME (p < 0.001). On the contrary, eyes with hard exudates showed less improvement in their response density at the end of the study (p < 0.001). We observed a significant relationship between BCVA and the response density achieved at the end of the study (p = 0.012). Eyes with severe and proliferative DR significantly shortened implicit time compared to those with moderate DR (p = 0.04).ConclusionsThe multifocal electroretinogram allowed us to differentiate groups of eyes with DME according to their electrophysiological profile, both initially and after being treated with RNBZ. Ranibizumab increased the response density in all DME types included in the study, with a maximum response in those eyes with spongiform type. Once treated with RNBZ, the macular electrophysiological activity improved in eyes that had a well-preserved ellipsoid zone and ELM. The presence of hard exudates was a limitation to the response density achieved at the foveola.

Non-adherence to eye care in people with diabetes

ObjectiveEvaluate individual factors that impact adherence to eye care follow-up in patients with diabetes.Design and methodsA 4-year retrospective chart review was conducted for 1968 patients with diabetes over age 40...

Prevalence of Diabetic Retinopathy and Blindness in Indonesian Adults With Type 2 Diabetes

To report the prevalence of diabetic retinopathy (DR) and DR-related blindness in an Indonesian population with type 2 diabetes. Design: Population-based cross-sectional study. Community health centers. We recruited 1184 people aged older than 30 years with type 2 diabetes residing in Jogjakarta, Indonesia. Multistage, clustered random sampling based on regencies and districts in Jogjakarta was used. Detailed interviews, general and eye examinations, and anthropometric measurement were performed. Disc- and macula-centered retinal photographs were taken to assess DR. The definition of DR followed a modified Airlie House classification system and was categorized into mild, moderate, and vision-threatening DR (VTDR). Prevalence and severity of DR. The median (range) age and diabetes duration of participants was 59 (52-65) and 4 (2-9) years. The prevalence of DR was 43.1% (95% confidence interval 39.6%-46.6%), with mild, moderate, and severe NPDR and PDR to be 9.41%, 7.46%, 11.1%, and 12.1%, respectively. The prevalence of VTDR was 26.3% (23.1%-29.5%). Longer diabetes duration, higher fasting glucose, presence of hypertension, and foot ulcers were associated with DR and VTDR. The prevalence of bilateral blindness was 4% and 7.7% in persons with DR and VTDR. This study reports a high prevalence of any DR and VTDR among Indonesian adults with type 2 diabetes in urban and rural areas: approximately 1 in 4 adults with diabetes had VTDR and 1 in 12 of those with VTDR was bilaterally blind, suggesting the need for appropriate screening and management of DR among the Indonesian population.

Accuracy and reliability of retinal photo grading for diabetic retinopathy: Remote graders from a developing country and standard retinal photo grader in Australia.

BackgroundTo evaluate the accuracy and reliability of fundus retinal photos graded by local graders in Bangladesh with those graded by an expert at the Centre for Eye Research Australia (CERA) in the context of mass scale diabetic retinopathy (DR) screening in Bangladesh.MethodsA population-based cross-sectional study of 3,104 adults identified 213 (7.2%) eligible patients with diabetes of age ≥ 40 years in 2012–2013. Retinal photographs were collected using a non-mydriatic digital fundus retinal camera and a two-field imaging protocol. The photos were graded by two remote graders (G1 and G2) who were trained by a retinal specialist (RS) in Bangladesh, by the RS himself, and by a Centre for Eye Research Australia (CERA) grader. The local graders up skilled their grading ability by comparing 30% of the photos graded by the CERA grader with their own grades. Learning from that exercise was applied to the remaining 70% of photos, which were re-graded. Reliability and accuracy of grading amongst the graders were reported using cross tabulation, inter- and intra-grader reliability, and with sensitivity and specificity.ResultsOf 122 eyes from 61 patients, the mild (R1) DR was estimated to be 14 to 25%, pre-proliferative (R2) DR 4–8%, and proliferative (R3) DR 0.8 to 1.6%, whereas 25%, 8%, 18%, and 15% were found to be ungradable by CERA, RS, G1, and G2, respectively. Of 8 (6.6%) eyes identified as R2 by the CERA grader, 5 (63%), 3 (38%) and 3 (38%) were correctly classified as R2, whereas the rest were classified either as R1 or R3 but none were classified as no DR (R0) or ungradable by the RS, G1 and G2, respectively. After getting experience reviewing the 30% test set graded by the CERA grader, the local graders graded moderate and severe DR with 100% accuracy. After excluding ungradable photos, the sensitivity (specificity) relative to the CERA grader was 82% (88%) before and 80% (93%) after training for G1 and 56% (87%) before and 77% (90%) after training for G2. In case of maculopathy, the CERA grader reported 11.2% eyes with maculopathy, which included 100% of the 4.9% by RS, 6.6% by G1, and 7.4% by G2.ConclusionsLocal graders in Bangladesh are able to grade retinal photos with high accuracy if the DR is at least of a moderate level. With appropriate training and experience, local graders have the ability to contribute significantly to the grading of millions of retinal photos, which required grading in resource- poor countries.

Highlights from the Diabetes UK Professional Conference 2017

Highlights from the Diabetes UK Professional Conference 2017

Hemorrhage and/or Microaneurysm Severity and Count in Ultrawide Field Images and Early Treatment Diabetic Retinopathy Study Photography

To evaluate detection of hemorrhage and/or microaneurysm (H/Ma) using ultrawide field (UWF) retinal imaging as compared with standard Early Treatment Diabetic Retinopathy Study (ETDRS) 7-field photographs (ETDRS photos). Single-site comparative study of UWF images and ETDRS photos. One hundred twenty-six eyes of 69 patients with no diabetic retinopathy (DR) or mild or moderate nonproliferative DR (NPDR). Stereoscopic 200° UWF images and stereoscopic 35mm 30° 7-field color photographs were acquired on the same visit. Images were graded for severity and distribution of H/Ma. H/Mas were counted in ETDRS fields 2 to 7 in both ETDRS photos and UWF images. H/Mas in the UWF peripheral fields were also counted. Kappa (κ) and weighted κ statistics for agreement. Number of H/Ma within and outside ETDRS fields identified in UWF images and ETDRS photos. Distribution of DR severity by ETDRS photos was 24 (19.0%) no DR, 48 (38.1%) mild NPDR, and 54 (42.9%) moderate NPDR. A total of 748 of 756 fields (98.9%) were gradable for H/Mas on ETDRS photos and UWF images. Simple κ/weighted κ statistics for severity of H/Ma: all fields 0.61/0.69, field 2 0.70/0.77, field 30.62/0.73, field 4 0.50/0.62, field 5 0.54/0.65, field 6 0.64/0.70, and field 7 0.58/0.63 with overall exact agreement in 81.3% and within 1 step in 97.9% of fields. A greater proportion of fields was graded a more severe H/Ma level in UWF images than in the corresponding ETDRS photos (UWF: 12.7% vs. 6.5%). Evaluating comparable areas in UWF images and ETDRS photos (fields 2-7), a mean of 42.8 H/Mas were identified usingETDRS photos and 48.8 in UWF images (P= 0.10). An additional mean of 21.3 H/Mas (49.8% increase, P<0.0001) were identified in the peripheral fields of the UWF images. There is good to excellent agreement between UWF images and ETDRS photos in determining H/Ma severity, with excellent correlation of H/Ma counts within ETDRS photo fields. UWF peripheral fields identified 49.8% more H/Ma, suggesting a more severe H/Ma in 12.7% of eyes. Given the additional lesions detected in peripheral fields and the known risks associated with H/Ma and peripheral lesions, quantification of H/Ma using UWF images may provide a more accurate representation of DR disease activity and potential greater accuracy in predicting DR progression.

EFFECT OF AN INTRAVITREAL DEXAMETHASONE IMPLANT ON DIABETIC MACULAR EDEMA AFTER CATARACT SURGERY.

To analyze the effects of a dexamethasone intravitreal implant (DEX; Ozurdex 700 μg; Allergan) administered immediately after cataract surgery in diabetic patients. This prospective, single-arm, open label study (NCT01748487 at ClinicalTrials.gov) involved Type 2 diabetic patients with at least mild diabetic retinopathy (DR) who underwent cataract surgery and DEX insertion after phacoemulsification, and intraocular lens implantation were enrolled. Best-corrected visual acuity and central retinal thickness (CRT) measured by spectral-domain optical coherence tomography were recorded at 1 week preoperatively, and 1 week, 1 month, and 3 months after surgery. Adverse events were also recorded. Twenty-four eyes of 24 patients (17 [70.8%] men; mean age 63.7 ± 8.7 years) with mild nonproliferative DR (41.7%), moderate nonproliferative DR (33.3%), severe nonproliferative DR (16.7%), or treated proliferative DR (8.3%) were selected. After DEX treatment, mean CRT changed from 241.1 μm (95% confidence interval, 227.5-254.6 μm) at baseline to 236.9 μm (95% confidence interval, 223.9-249.9 μm) at 1 week (P = 0.09), 238.9 μm (95% confidence interval, 225.5-252.3 μm) at 1 month (P = 0.44), and 248 μm (95% confidence interval, 232.4-260.8 μm) at 3 months (P = 0.15). No eyes showed a postoperative increase >50 μm in the CRT at any visit. A 10% increase in CRT was found in 8.3% of eyes. Mean best-corrected visual acuity significantly improved from 0.37 (20/50) at baseline to 0.19 (20/30) at 1 week, 0.12 (20/25) at 1 month, and 0.12 (20/25) at 3 months (P < 0.001 for each comparison). Mean intraocular pressure before surgery was 13.8 mmHg, and none of the patients developed an intraocular pressure ≥22 mmHg at any visit. None of the patients developed any serious adverse events during the follow-up. These short-term results suggest that a single DEX injection intraoperatively after phacoemulsification could avoid an increase in CRT after cataract surgery in diabetic patients.

En Face Doppler Optical Coherence Tomography Measurement of Total Retinal Blood Flow in Diabetic Retinopathy and Diabetic Macular Edema

Alterations in ocular blood flow play an important role in the pathogenesis and progression of diabetic retinopathy (DR). However, the measurement of retinal blood flow in clinical studies has been challenging. En face Doppler optical coherence tomography (OCT) provides an effective method for measuring total retinal blood flow (TRBF) in the clinic. To investigate TRBF in eyes with DR of varying severity, with or without diabetic macular edema (DME), using en face Doppler OCT. This was a cross-sectional study conducted from May 23, 2014, to January 11, 2016, which analyzed 41 eyes with DR from 31 diabetic patients, 20 eyes without DR from 11 diabetic patients, and 16 eyes from 12 healthy age-matched controls, all at the New England Eye Center in Boston, Massachusetts. Participants were imaged with a high-speed, swept-source OCT prototype at 1050-nm wavelength using repeated en face Doppler OCT raster scans, comprising 600 × 80 axial scans and covering a 1.5 × 2-mm2 area centered at the optic disc. The TRBF was automatically calculated using custom Matlab software. This study included 41 eyes with DR from 31 diabetic patients (mean [SD] age, 62.8 [13.4] years; 12 were female patients), 20 eyes without DR from 11 diabetic patients (mean [SD] age, 58.8 [10.1] years; 5 were female patients), and 16 eyes from 12 healthy age-matched controls (mean [SD] age, 57.9 [8.1] years; 8 were female participants). The mean (SD) TRBF was 28.0 (8.5) µL/min in the eyes with DME, 48.8 (13.4) µL/min in the eyes with DR but without DME, 40.1 (7.7) µL/min in the diabetic eyes without retinopathy, and 44.4 (8.3) µL/min in age-matched healthy eyes. A difference in TRBF between the eyes with DME that were treated and the eyes with DME that were not treated was not identified. The TRBF was consistently low in the eyes with DME regardless of DR severity. The eyes with moderate nonproliferative DR but without DME exhibited a wide range of TRBF from 31.1 to 75.0 µL/min, with the distribution being highly skewed. High-speed en face Doppler OCT can measure TRBF in healthy and diabetic eyes. Diabetic eyes with DME exhibited lower TRBF than healthy eyes (P ≤ .001). Further longitudinal studies of TRBF in eyes with DR would be helpful to determine whether reduced TRBF is a risk factor for DME.

Stronger relationship of serum apolipoprotein A-1 and B with diabetic retinopathy than traditional lipids

Aim:Diabetic retinopathy (DR) is the most common preventable cause of blindness where early detection and treatment can be sight-saving. Search for biomarkers of the disease has been relentless. We aimed to determine whether lipoproteins apolipoproteins A1 and B1 (Apo-A1 and Apo-B1) have stronger associations with DR in contrast to conventionally measured low-density lipoprotein (LDL) and high-density lipoprotein cholesterol levels.Materials and Methods:We performed a cross-sectional study and studied 117 patients. Serum lipid profile was assessed by autoanalyzer. Serum Apo-A1 and Apo-B were measured using immunoturbidimetric kit on an autoanalyzer. Apo-B/A1 ratio was calculated. Retinopathy was graded from the digital retinal photographs, taken with nonmydriatic auto fundus camera and classified according to International Clinical DR Disease Severity Scale.Results:Mean Apo-A1 for mild, moderate, severe retinopathy, and proliferative DR (PDR) shows a significant negative correlation (P = 0.001) with severity of retinopathy. Mean Apo-B for mild, moderate, severe, PDR displayed a significant positive correlation with severity of retinopathy (P = 0.001). Mean Apo-B/A1 for mild, moderate, severe, PDR showed highly significant positive correlation with severity of retinopathy (P < 0.001). In contrast, mean LDL for mild, moderate, severe, PDR showed insignificant association with severity of DR (P = 0.081).Conclusion:Apo-A1 and Apo-B have a stronger association with the development of DR than traditional lipids and can thus facilitate early detection and treatment of the disease.

Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs

Deep learning is a family of computational methods that allow an algorithm to program itself by learning from a large set of examples that demonstrate the desired behavior, removing the need to specify rules explicitly. Application of these methods to medical imaging requires further assessment and validation. To apply deep learning to create an algorithm for automated detection of diabetic retinopathy and diabetic macular edema in retinal fundus photographs. A specific type of neural network optimized for image classification called a deep convolutional neural network was trained using a retrospective development data set of 128 175 retinal images, which were graded 3 to 7 times for diabetic retinopathy, diabetic macular edema, and image gradability by a panel of 54 US licensed ophthalmologists and ophthalmology senior residents between May and December 2015. The resultant algorithm was validated in January and February 2016 using 2 separate data sets, both graded by at least 7 US board-certified ophthalmologists with high intragrader consistency. Deep learning-trained algorithm. The sensitivity and specificity of the algorithm for detecting referable diabetic retinopathy (RDR), defined as moderate and worse diabetic retinopathy, referable diabetic macular edema, or both, were generated based on the reference standard of the majority decision of the ophthalmologist panel. The algorithm was evaluated at 2 operating points selected from the development set, one selected for high specificity and another for high sensitivity. The EyePACS-1 data set consisted of 9963 images from 4997 patients (mean age, 54.4 years; 62.2% women; prevalence of RDR, 683/8878 fully gradable images [7.8%]); the Messidor-2 data set had 1748 images from 874 patients (mean age, 57.6 years; 42.6% women; prevalence of RDR, 254/1745 fully gradable images [14.6%]). For detecting RDR, the algorithm had an area under the receiver operating curve of 0.991 (95% CI, 0.988-0.993) for EyePACS-1 and 0.990 (95% CI, 0.986-0.995) for Messidor-2. Using the first operating cut point with high specificity, for EyePACS-1, the sensitivity was 90.3% (95% CI, 87.5%-92.7%) and the specificity was 98.1% (95% CI, 97.8%-98.5%). For Messidor-2, the sensitivity was 87.0% (95% CI, 81.1%-91.0%) and the specificity was 98.5% (95% CI, 97.7%-99.1%). Using a second operating point with high sensitivity in the development set, for EyePACS-1 the sensitivity was 97.5% and specificity was 93.4% and for Messidor-2 the sensitivity was 96.1% and specificity was 93.9%. In this evaluation of retinal fundus photographs from adults with diabetes, an algorithm based on deep machine learning had high sensitivity and specificity for detecting referable diabetic retinopathy. Further research is necessary to determine the feasibility of applying this algorithm in the clinical setting and to determine whether use of the algorithm could lead to improved care and outcomes compared with current ophthalmologic assessment.

Visualization of 3 Distinct Retinal Plexuses by Projection-Resolved Optical Coherence Tomography Angiography in Diabetic Retinopathy

Projection artifacts in optical coherence tomography angiography (OCTA) blur the retinal vascular plexuses together and limit visualization of the individual plexuses. To describe projection-resolved (PR) OCTA in eyes with diabetic retinopathy (DR) and healthy eyes. In this case-control study, patients with DR and healthy controls were enrolled in this observational study from January 26, 2015, to December 4, 2015, at a tertiary academic center. Spectral-domain, 70-kHz OCT obtained 3 × 3-mm macular scans. The PR algorithm suppressed projection artifacts. A semiautomated segmentation algorithm divided PR-OCTA into superficial, intermediate, and deep retinal plexuses. Two masked graders examined 3-layer PR-OCTA and combined angiograms for nonperfusion and abnormal capillaries. Retinal nonperfusion and capillary abnormalities and the diagnostic accuracy of detecting DR. Twenty-nine eyes of 15 healthy individuals (mean [SD] age, 36.2 [13.4] years; 11 women) and 47 eyes of 29 patients with DR (mean [SD] age, 55.5 [11.9]; 10 women) underwent imaging. PR-OCTA revealed 3 distinct retinal plexuses in their known anatomical locations in all eyes. The intermediate and deep plexuses of healthy eyes revealed capillary networks of uniform density and caliber, whereas the superficial plexus revealed vessels in the familiar centripetal branching pattern. In eyes with DR, 3-layer PR-OCTA disclosed incongruent areas of nonperfusion and varied vessel caliber and density in the deeper plexuses. Masked grading of capillary nonperfusion on 3-layer PR-OCTA detected DR with 100% sensitivity (95% CI, 90.8%-100%) and 100% specificity (95% CI, 85.4%-100%). With unsegmented retinal angiograms, the sensitivity and specificity were 78.7% (95% CI, 63.9%-88.8%) and 100% (95% CI, 85.4%-100%), respectively (P = .002 for sensitivity). On 3-layer PR-OCTA, sensitivity was 72.2% (95% CI, 54.6%-85.2%) for severe nonproliferative DR and proliferative DR eyes with generalized nonperfusion in 2 or more individual plexuses, but on combined angiogram, sensitivity was 25.0% (95% CI, 12.7%-42.5%) for generalized nonperfusion (P < .001). PR-OCTA disclosed dilated vessels in the intermediate and deep plexuses in 23 eyes (100%) with proliferative DR, 13 eyes (100%) with severe nonproliferative DR, 8 eyes (73%) with mild to moderate nonproliferative DR, and 0 control eyes. By presenting 3 retinal vascular plexuses distinctly, PR-OCTA reveals capillary abnormalities in deeper layers with clarity and may distinguish DR from healthy eyes and severe DR from mild DR with greater accuracy compared with conventional OCTA.