Comprehensive Modality Integration for Diabetic Retinopathy Image Analysis

We propose an algorithm for the detection of retinal landmarks (optic nerve head or optic disc, macula, and vasculature) based on optic cup location and anatomical structural details from diabetic retinopathy (DR) images of both left and right eye. Our algorithm uses color fundus images obtained from mydriatic camera. The algorithm proceeds through four main steps 1. Color image pre-processing- to enhance and remove noise from the image. 2. Detection of optic nerve head-The optic nerve head is located using a fact that the optic cup is the brightest region in the optic nerve head. At the same time exudates (DR lesion) which appear in same gray level as optic nerve head are suppressed since we only concentrate on optic cup and optic nerve head. So by calculating the mean value of the intensities of 50 ? 50 subimages (50 ? 50 is approximate area of optic cup) throughout the image and then selecting that 50 ? 50 sub image with the highest mean value, locates optic cup in the image. Using this, optic nerve head is located by increasing the area of interest around the optic cup. Since we detect optic cup first, which is embedded in optic nerve head there will not be any false detection of optic nerve head when size and shape of the exudates (DR lesion) are same as that of optic nerve head 3.Detection of macula-It is located at a distance of approximately twice the diameter of the optic nerve head just below the horizontal axis of the optic nerve head. 4. Detection of vasculature-We have used logical AND operation on two images, one being a thresholded image and another being an edge detected image. The thresholding is done on an adaptive histogram equalized image. Edge detection is done using canny edge detector. Proposed algorithm has been tested on both normal and DR images. Detected optic disc area is validated by comparing it with expert ophthalmologists' hand-drawn ground-truths. The quantitative performance is evaluated by calculating sensitivity, specificity and predictive value. Overall sensitivity (Se), specificity (Sp) and predictive value (PV) obtained in detecting optic nerve head from normal images and from abnormal images are 97.2%, 99.72%, and 88.75% and 93.93%, 99.72%, and 84.18% respectively.

Artificial intelligence for diabetic retinopathy screening in Africa

목적: 당뇨병 환자와 정상인에서 스펙트럼 영역 빛간섭 단층촬영(Carl Zeiss Meditec, Dublin, CA)을 통한 신경절 세포-내망상층 및망막신경섬유층 두께의 차이에 대하여 분석한다. 대상과 방법: 정상안 92안과 당뇨안 42안을 대상으로 스펙트럼 영역 빛간섭단층촬영을 시행하여 망막신경섬유층 두께와 신경절 세포-내망상층 두께의 차이를 비교 분석하였다. 결과: 정상군 92명, 당뇨망막병증이 없는 당뇨환자 22명, 당뇨망막병증이 있는 당뇨환자 26명으로 각 군을 나누어 분석한 결과 당뇨유무나 당뇨망막병증 유무는 망막신경섬유층 두께 차이에 영향을 주지 않았다. 신경절 세포-내망상층 두께는 당뇨가 있는 환자에서,그리고 당뇨망막병증이 있는 환자에서 더욱 얇아지는 경향이 있었으며 특히 위쪽 망막 구역에서 정상군 82.24 ± 7.21 μm, 당뇨망막병증이 없는 당뇨환자군 81.86 ± 9.53 μm, 당뇨망막병증이 있는 당뇨환자군 76.77 ± 14.13 μm로 유의하게 감소하였다(p=0.029). 결론: 당뇨와 당뇨망막병증에 의한 영향에 의해 망막 신경절 세포의 두께 감소가 유발되며 이러한 변화는 특히 위쪽 망막에서 먼저일어나는 것을 확인하였다.

To quantitatively describe and evaluate a new image processing technique for estimating the Foveal Avascular Zone (FAZ) in subjects with Diabetic Retinopathy and myopes. From a total of 328 images obtained from Diabetic Retinopathy (113), myopes (120) and normal (93), the FAZ dimensions were quantified using a new image processing algorithm. These parameters were also determined manually and by the OCT manufacturer's inbuilt algorithm. In the new technique, the images were first pre-processed by using a DOG filter iteratively before being complemented followed by a Prewitt edge detection and repeated image dilation at angles of 00, 450 and 900. Image closure was then applied followed by noise and small object removal which resulted in the segmented boundary. For deeper insight into shape change, in addition to the diameter of the FAZ other parameters such as the area, diameter, major axis, minor axis, orientation, perimeter vessel avascular density (VAD), Vessel diameter Index (VDI), etc. were obtained. The circularity index was calculated using the FAZ area and perimeter parameters. The mean FAZ diameter (mm) by the new automated technique, manual-segmentation (ground truth), and inbuilt instrument algorithm were 0.67 ± 0.87, 0.67 ± 0.72 and 0.61 ± 0.14. The mean of FAZ area (mm2) was 0.36 ± 0.10, 0.33 ± 0.09 and 0.43 ± 0.14 in normal, myopia and diabetic subjects respectively. The new technique shows considerable improvement in accuracy (mean ± SD) when compared to the inbuilt system segmentation and the ground truth (manual marking by an expert clinician). The study results show that the FAZ area in Diabetic Retinopathy is significantly different (p=0.003) when compared to myopic eyes (p=0.016) and normals.

Diabetic macular oedema (DMO) results from alterations of several biochemical pathways in diabetic eyes. Centre-involving DMO is an important cause of visual loss in diabetes. Anti-vascular endothelial growth factor agents are now the mainstay of centre-involving DMO treatment. Oedema that does not achieve optimal response to these agents occurs in a sizeable proportion of eyes and is called refractory or persistent DMO. Management of refractory DMO is challenging. In this paper, the pathophysiology of DMO, and the definitions used in various studies are summarised. Therapeutic options for refractory DMO management including corticosteroids, laser, combination therapies, and surgery are explored. Novel agents on the horizon for DMO control that are being investigated at present are discussed as well. A literature review was performed and a summary of the research studies for each of the agents is provided in order to guide the reader regarding the existing evidence for their application in DMO. Importance of early recognition of disease and prompt treatment to achieve best visual outcome is discussed. Utility of optical coherence tomography to guide disease diagnosis and monitoring is highlighted. An algorithmic approach for DMO management is described. Finally, the impact that personalized medicine and genetics might have on DMO management is assessed.

Background: To evaluate the earliest retinal morphological and functional changes in diabetic eyes without or with early signs of diabetic retinopathy (DR). Methods: Twenty-two eyes with no DR (noDR group), 22 eyes with mild DR (DR group), and 18 healthy nondiabetic eyes (controls) were enrolled. All eyes were studied by means of spectral domain optical coherence tomography (OCT), OCT angiography (OCTA), and multifocal electroretinogram (mfERG). Results: A significantly higher number of OCT hyperreflective intraretinal foci (HRF) was found in both noDR and DR groups versus controls, but not between DR groups. The OCTA parameters of the superficial vascular plexus (SVP) were significantly reduced in the noDR group both versus controls and DR group (p < 0.05). The OCTA parameters of the intermediate capillary plexus (ICP) were significantly reduced in the DR group versus controls. An increased number of altered hexagons on mfERG was found in the noDR versus the DR group (p = 0.0192). Conclusions: Retinal vascular and functional parameters are differently involved in diabetic eyes; major vascular changes in the SVP and functional alterations of the mfERG are present in diabetic eyes with no clinical microvascular signs of DR, while ICP is mainly involved when early ophthalmoscopic signs of DR are present. The integrated use of mfERG and OCTA provides new significant insights into the pathogenesis of diabetic related retinal disease.

Diabetic retinopathy (DR) is an eye disease which is caused by high blood sugar and high blood pressure and damage the blood vessels in the back (retina) of the eye. Diabetic retinopathy (DR) is not a reversible process and treatment only sustain vision. The number of ophthalmologists cannot meet the growing demands around the world. This study focuses on the automatic diagnosis of the disease through deep learning. Convolutional neural network (CNN) is more widely used as a deep learning method in medical image analysis. Multiclass image classification of images into non referable DR and referable DR has been done using proposed Convolutional neural network (CNN) model. For multiclass classification problem, the sensitivity, specificity and F2 score value for class 0 (no DR) are 81.75 %, 91.06 % and 81.80 % respectively. Whereas for class 1 (non-severe DR), sensitivity, specificity and F2 score values are 71.28 %, 78.52 % and 70.01 % respectively. Similarly for class 2 (sever DR), sensitivity, specificity and F2 score values are 73.03 %, 93.39 % and 75.08 % respectively. Bayesian optimization has been performed for tuning learning rate and gives optimal learning rate 0.000358 through the optimization process. The customized CNN is then trained using 0.000358 learning rate and then tested on test data The images in dataset have poor contrast and consists of impulse noises. Contrast limited adaptive histogram equalization (CLAHE) method is used to improve the contrast of the image followed by median filter to remove noise present in DR image.

Diabetic retinopathy (DR) refers to a diabetes complexity that immensely impacts the eyes. This is classified into 5 various stages of the severity in accordance with the international convention. Despite that, optimization of a grading model to have a robust generalizability needs a huge number of balanced training data that is very complicated to gather, especially for greater levels of severity. A vast amount of medical data is complex and has a very high-priced method which requires cooperation between the clinics and researchers. The issue is usually attempted to be figured out with the usage of the traditional methods of data augmentation by making certain changes to images of retina dataset for instance rotation, cropping, size and zooming. In this suggested paper, the latest methods or techniques of data augmentation is exhibited which is called as deep convolutional generative adversial network (DC-GAN) and variational auto encoders (VAE). This is a particular method which is responsible for the production of artificial medical images. In addition to this, to improve DR, we can also take the aid of the classification models which are resnet50, densenet201, InceptionV3 and VGG19 for the purpose of classification of the eye related diseases. The proposed method is depicted on the Asia Pacific Tele-Ophthalmology Society (APTOS)-Blindness dataset. First, the present-day online data augmentation techniques have been utilized, and the artificial images of retina are produced by the ease of DCGAN. Then, a method of classifying is used for both techniques. Ultimately, after the method training which is done by using the real &amp; synthetic clinical images and the outcome exhibits which the proposed model determines every stage or phase of DR and achieve the accuracy of 98.66% with using of ResNet-50 which is contrary to the current existing techniques.

ObjectiveDiabetic retinopathy (DR), a major diabetes complication, is the fifth leading cause of global blindness and moderate-to-severe visual impairment. DR is categorized into non-proliferative and proliferative forms, with the latter involving neovascularization. Adaptive optics (AO) imaging provides high-resolution, in vivo visualization of retinal microstructures, offering enhanced insight into disease pathology.MethodsA comprehensive PubMed search was conducted for studies published before August 1, 2025, using the keywords “diabetic retinopathy” and “adaptive optics”. The review focused on studies assessing cone photoreceptors, retinal vasculature, and clinical signs using AO imaging.ResultsForty-two studies were included. Many examined cone photoreceptor integrity across DR stages, with or without diabetic macular edema (DME). AO imaging revealed microstructural changes, enabling the correlation of cone metrics with other imaging modalities, including optical coherence tomography angiography, electroretinography, microperimetry, and color vision tests. Several studies analyzed capillary-level vascular alterations and their association with DR severity and DME. Some works have integrated AO data with blood flow assessments using techniques such as laser speckle flowgraphy and AO scanning laser ophthalmoscopy. Lastly, several studies have identified clinical signs related to DR and described their features on AO imaging.DiscussionAdaptive optics imaging reveals early and progressive alterations in photoreceptors, retinal microvasculature, and clinical signs such as microaneurysms and hard exudates in diabetic retinopathy, supporting its role as a valuable tool for understanding disease mechanisms and guiding early diagnosis and monitoring.ConclusionsAdaptive optics is a valuable imaging tool that improves understanding of retinal microarchitecture in diabetic retinopathy. It aids diagnosis, monitoring, and prognosis by enabling detailed evaluation of cellular and vascular alterations.

OBJECTIVE To compare efficiency of nonmydriatic ultrawide field retinal imaging (UWFI) and nonmydriatic fundus photography (NMFP) in a diabetic retinopathy (DR) ocular telehealth program. RESEARCH DESIGN AND METHODS Patients in this retrospective, comparative cohort study underwent NMFP and UWFI between 1 November 2011 and 1 November 2012. Images were evaluated for DR and diabetic macular edema (DME) by certified graders using a standard protocol at a centralized reading center. Identification of DR, image evaluation time, and rate of ungradable eyes were compared. RESULTS NMFP and UWFI were performed in 1,633 and 2,170 consecutive patients, respectively. No statistically significant differences were found between groups regarding age, diabetes duration, sex, ethnicity, or insulin use. The ungradable rate per patient for DR (2.9 vs. 9.9%, P < 0.0001) and DME (3.8 vs. 8.8%, P < 0.0001) was lower with UWFI than with NMFP. With UWFI, the median image evaluation time per patient was reduced from 12.8 to 9.2 min (P < 0.0001). The identification of patients with DR (38.4 vs. 33.8%) and vision-threatening DR (14.5 vs. 11.9%) was increased with UWFI versus NMFP. In a consecutive subgroup of 502 eyes of 301 patients with DR, the distribution of peripheral retinal lesions outside Early Treatment Diabetic Retinopathy Study fields suggested a more severe DR level in 9.0% (45 eyes). CONCLUSIONS In a standardized DR ocular telehealth program, nonmydriatic UWFI reduced the ungradable rate by 71% (to <3%) and reduced image evaluation time by 28%. DR was identified 17% more frequently after UWFI, and DR peripheral lesions suggested a more severe DR level in 9%. These data suggest that UWFI may improve efficiency of ocular telehealth programs evaluating DR and DME.

Diabetic retinopathy (DR) is a microvascular complication of diabetes with a high incidence and one of the four major blind eye diseases recognized in the world. At present, in clinical practice, the analysis and diagnosis of fundus retinal photos are mainly carried out by professional doctors, but this process is very dependent on the clinical experience of doctors, which is time-consuming and labor-intensive, misdiagnosis. Based on this, this paper proposes a DR automatic classification method that integrates attention mechanism and EfficientNet to improve the accuracy of classification and discrimination. In view of the problems in the DR data set, processing such as elimination, amplification and normalization is performed; The attention mechanism is integrated into the EfficientNet network, the DR data set is used to learn and train the improved EfficientNet., and the features are extracted. At the same time, the hole convolution is added to increase the receptive field, and the DR images are divided into 5 categories according to their levels. The method proposed in this paper achieves good results in the DR image classification task, with accuracy, precision, sensitivity and F1 score of 0.857, 0.838, 0.886 and 0.767, respectively. The results show that the DR classification algorithm proposed in this paper has good classification performance.

Our website uses cookies to enhance your experience. By continuing to use our site, or clicking "Continue," you are agreeing to our Cookie Policy | Continue JAMA Ophthalmology HomeNew OnlineCurrent IssueFor Authors Podcast Publications JAMA JAMA Network Open JAMA Cardiology JAMA Dermatology JAMA Health Forum JAMA Internal Medicine JAMA Neurology JAMA Oncology JAMA Ophthalmology JAMA Otolaryngology–Head & Neck Surgery JAMA Pediatrics JAMA Psychiatry JAMA Surgery Archives of Neurology & Psychiatry (1919-1959) JN Learning / CMESubscribeJobsInstitutions / LibrariansReprints & Permissions Terms of Use | Privacy Policy | Accessibility Statement 2023 American Medical Association. All Rights Reserved Search All JAMA JAMA Network Open JAMA Cardiology JAMA Dermatology JAMA Forum Archive JAMA Health Forum JAMA Internal Medicine JAMA Neurology JAMA Oncology JAMA Ophthalmology JAMA Otolaryngology–Head & Neck Surgery JAMA Pediatrics JAMA Psychiatry JAMA Surgery Archives of Neurology & Psychiatry Input Search Term Sign In Individual Sign In Sign inCreate an Account Access through your institution Sign In Purchase Options: Buy this article Rent this article Subscribe to the JAMA Ophthalmology journal

Using Deep Learning Models to Characterize Major Retinal Features on Color Fundus Photographs

Diabetic retinopathy (DR) is a common eye disease that people get from diabetes. About 33.7% of the people with diabetes have DR. With our datas, which are pictures of the eyeball with and without DR, we tried different convolutional neural network (CNN) models to get the best accuracy score. We tested our datas with a default CNN model, and 5 different pre-trained models: MobileNet, VGG16, VGG19, Inception V3, and Inception ResNet V2. The default CNN model didn’t perform very well, getting only 10.4%. The pre-trained model also didn’t perform as good as expected, so we decided to use GRU with the models, which increases the score. For the higher accuracy, we added bidirectional GRU to train the whole parameters in the model. The 5 different pre-trained models scored an average of 74.2% accuracy score, and Inception ResNet V2 with bidirectional GRU included scored the highest accuracy, achieving 83.57%. For additional study, we used a class activation map to spot the abnormal parts of the eyeball with DR, and we could spot abnormal veins and bleeding on the eyeball. However, our research has limitations on that we did not use the segmentation methods, which is more advanced technique compared to classification, such as U-net, Fully Convolutional Network (FCN), Deep Lab V3, and Feature Pyramid Network. Furthermore, even though our model classified 5 different classes, the fact that the highest accuracy score was lower than 90% is also a limitation. For further study, we would prepare a masking method for applying segmentation methods to our dataset.

Loss of the inner blood-retinal barrier (BRB) integrity is a main feature of ocular diseases such as diabetic macular edema. However, there is a lack of clarity on how inner BRB function is modulated within the diabetic retina. The current study examined whether eucalyptol inhibited inner BRB destruction and aberrant retinal angiogenesis in 33 mM glucose-exposed human retinal microvascular endothelial (RVE) cells and db/db mice. This study further examined the molecular mechanisms underlying endothelial dysfunction including retinal endoplasmic reticulum (ER) stress and angiopoietin (Ang)/Tie axis in conjunction with vascular endothelial growth factor (VEGF). Eucalyptol is a naturally occurring monoterpenoid and an achiral aromatic component of many plants including eucalyptus leaves. Nontoxic eucalyptol reduced the production of amyloid-β (Aβ) protein in glucose-loaded RVE cells and in diabetic mice. This natural compound blocked apoptosis of Aβ-exposed RVE cells in diabetic mouse eyes by targeting ER stress via the inhibition of PERK-eIF2α-ATF4-CHOP signaling. Eucalyptol promoted activation of the Ang-1/Tie-2 pathway and dual inhibition of Ang-2/VEGF in Aβ-exposed RVE cells and in diabetic eyes. Supply of eucalyptol reversed the induction of junction proteins in glucose/Aβ-exposed RVE cells within the retina and reduced permeability. In addition, oral administration of eucalyptol reduced vascular leaks in diabetic retinal vessels. Taken together, these findings clearly show that eucalyptol inhibits glucose-induced Aβ-mediated ER stress and manipulates Ang signaling in diabetic retinal vessels, which ultimately blocks abnormal angiogenesis and loss of inner BRB integrity. Therefore, eucalyptol provides new treatment strategies for diabetes-associated RVE defects through modulating diverse therapeutic targets including ER stress, Ang-1/Tie-2 signaling, and Ang-2/VEGF.