This dataset is a large set of high-resolution retina images taken under a variety of imaging conditions. A left and right field is provided for every subject. Images are labeled with a subject id as well as either left or right (e.g. 1_left.jpeg is the left eye of patient id 1). A clinician has rated the presence of diabetic retinopathy in each image on a scale of 0 to 4, according to the following scale: 0 - No DR 1 - Mild 2 - Moderate 3 - Severe 4 - Proliferative DR Total Images: 35126. The distribution of labels is: 0: 25810, 1: 2443, 2: 5292, 4: 708, 3: 873 Your task is to create an automated analysis system capable of assigning a score based on this scale. The images in the dataset come from different models and types of cameras, which can affect the visual appearance of left vs. right. Some images are shown as one would see the retina anatomically (macula on the left, optic nerve on the right for the right eye). Others are shown as one would see through a microscope cond

Diabetic retinopathy is the leading cause of blindness in the working-age population of the developed world. It is estimated to affect over 93 million people.

retina

The US Center for Disease Control and Prevention estimates that 29.1 million people in the US have diabetes and the World Health Organization estimates that 347 million people have the disease worldwide. Diabetic Retinopathy (DR) is an eye disease associated with long-standing diabetes. Around 40% to 45% of Americans with diabetes have some stage of the disease. Progression to vision impairment can be slowed or averted if DR is detected in time, however this can be difficult as the disease often shows few symptoms until it is too late to provide effective treatment.

Currently, detecting DR is a time-consuming and manual process that requires a trained clinician to examine and evaluate digital color fundus photographs of the retina. By the time human readers submit their reviews, often a day or two later, the delayed results lead to lost follow up, miscommunication, and delayed treatment.

Clinicians can identify DR by the presence of lesions associated with the vascular abnormalities caused by the disease. While this approach is effective, its resource demands are high. The expertise and equipment required are often lacking in areas where the rate of diabetes in local populations is high and DR detection is most needed. As the number of individuals with diabetes continues to grow, the infrastructure needed to prevent blindness due to DR will become even more insufficient.

The need for a comprehensive and automated method of DR screening has long been recognized, and previous efforts have made good progress using image classification, pattern recognition, and machine learning. With color fundus photography as input, the goal of this competition is to push an automated detection system to the limit of what is possible – ideally resulting in models with realistic clinical potential. The winning models will be open sourced to maximize the impact such a model can have on improving DR detection.

Acknowledgements This competition is sponsored by the California Healthcare Foundation.

Retinal images were provided by EyePACS, a free platform for retinopathy screening.

Diabetic Retinopathy Detection Competition Dataset img size/shape

Dataset

This dataset was created by Guido Zuidhof

Released under Data files © Original Authors

Contents

Dataset Card for Dataset Name

All the images of the dataset come from this kaggle dataset. Some minor modifications have been made to the metadata. All credit goes to the original authors and the contributor on Kaggle.

  Dataset Details





  Dataset Description

The EyePACS dataset consists of retinal images originally published in the Kaggle competition "Diabetic Retinopathy Detection". This version includes a subset of the original data, specifically the… See the full description on the dataset page: https://huggingface.co/datasets/bumbledeep/eyepacs.

Training versions of the AlexNet Convolutional Neural Network (CNN), related to the evaluation of eye diseases. -The following datasets were used in this work: LAG. Contains fundus images with positive and negative glaucoma samples obtained from Beijing Tongren Hospital. Each fundus image is diagnosed by qualified glaucoma specialists, taking into consideration of both morphologic and functional analysis. https://arxiv.org/abs/1903.10831 APTOS. Contains images of diabetic retinopathy that were used in the APTOS 2019 blindness screening competitions. Each image has been resized and cropped to have a maximum size of 1024px. A certified clinician rated each image according to the severity of diabetic retinopathy on a scale of 0 to 4. https://www.kaggle.com/c7934597/resized-2015-2019-diabetic-retinopathy-detection/metadata/ HRF. Contains 15 images of healthy patients, 15 images of patients with diabetic retinopathy and 15 images of glaucomatous patients. They were captured by a Canon CR-1 fundus camera with a field of view of 45 degrees with a resolution of 3504×2336 px. https://www5.cs.fau.de/research/data/fundus-images/ ODIR. Contains images of 5000 patients with various eye diseases collected by Shanggong Medical Technology Co., Ltd. from different hospitals/medical centers in China. The fundus images are captured with various cameras on the market, resulting in varied image resolutions. They classify patients into eight labels based on the images of both eyes. https://odir2019.grand-challenge.org/dataset/ sjchoi86-HRF. This dataset contains 601 fundus images divided into 4 groups: normal (300 images), glaucoma (101 images), cataract (100 images) and retina disease (100 images). https://github.com/cvblab/retina_dataset -Specifications of the data used for training in each version of the Convolutional Neural Network (CNN): netTransfer: Based on glaucoma and non-glaucoma imag-es cases existing in the LAG-database. netTransfer2: Based on glaucoma and non-glaucoma imag-es cases existing in the LAG-database and the sjchoi86-HRF database (this data-base also contains images on other pathologies, but only the glaucoma samples were used for this training) netTransfer3: Based on glaucoma, diabetic retinopathy and non-disease images cases existing in the LAG-database, sjchoi86-HRF database and the HRF database. netTransfer4: Based on glaucoma, diabetic retinopathy and non-disease images cases existing in the LAG-database, sjchoi86-HRF database, HRF database and the APTOS database. However, data from the fourth iteration was corrupted and then there for lost. netTransfer5: Based on glaucoma, diabetic retinopathy and non-disease images cases existing in the LAG-database, sjchoi86-HRF database, HRF database and the APTOS database (Images were cropped slightly before training). netTransfer6: Based on glaucoma, diabetic retinopa-thy and non-disease images cases existing in the LAG-database, sjchoi86-HRF database, HRF database, APTOS database and ODIR database (this database also contains images on several other diseases, but only the glaucoma and diabetic retinopathy samples were used for this training; furthermore, cropping was used for the images in order the eliminate black borders). PLEASE CITATE AS: Arias-Serrano I, Velásquez-López PA, Avila-Briones LN et al. Artificial intelligence based glaucoma and diabetic retinopathy detection using MATLAB — retrained AlexNet convolutional neural network [version 1; peer review: 1 approved with reservations]. F1000Research 2023, 12:14 (https://doi.org/10.12688/f1000research.122288.1)

Doenças associados ao diabetes são recorrentes. A retinopatia diabética interfere na visão do paciente e pode ocasionar o rompimento de vasos sanguíneos dos olhos, levando a cegueira caso não seja diagnosticada e tratada em tempo hábil. O diagnóstico pode ser feito através da análise de imagens de fundo de olho. A classificação entre imagens com lesão e não lesão pode auxiliar nesse processo. Foram extraídos dessas imagens os descritores de Histograma (H), Comprimento de Corrida (CC) e Matriz de Co-ocorrência (CO) e analisados os métodos K- vizinhos Próximos (K-NN), Máquinas de Vetores de Suporte (SVM) Linear e Radial, Árvore de Decisão (DT), Florestas Aleatórias (RF), Redes Neurais Artificiais (RNA), Análise de discriminante linear (LDA) e AdaBoost DT. O método LDA apresentou o melhor resultado nos experimentos realizado com 77,67% de acurácia. Methods Image data and labels were collected from Kaggle contest: https://www.kaggle.com/c/diabetic-retinopathy-detection/. After, 600 high resolution images were processed to extract Histogram, Co-occurence matrix and run length image descriptors and stored as CSV file available in this document.

Diabetic retinopathy (DR) is an eye disease triggered due to diabetes, which may lead to blindness. To prevent diabetic patients from becoming blind, early diagnosis and accurate detection of DR are vital. Deep learning models, such as convolutional neural networks (CNNs), are largely used in DR detection through the classification of blood vessel pixels from the remaining pixels. In this paper, an improved activation function was proposed for diagnosing DR from fundus images that automatically reduces loss and processing time. The DIARETDB0, DRIVE, CHASE, and Kaggle datasets were used to train and test the enhanced activation function in the different CNN models. The ResNet-152 model has the highest accuracy of 99.41% with the Kaggle dataset. This enhanced activation function is suitable for DR diagnosis from retinal fundus images.

This is a subset of the Kaggle retina fundus image data set. We select 1000 normal and 1000 Diabetic Retinopathy images, to make it smaller for the educational purpose. Original grade 0 is now labeled as "normal", and original grade 1-4 is not "dr".

Diabetic Retinopathy is the most prevalent cause of avoidable vision impairment, mainly affecting working age population in the world. Recent research has given a better understanding of requirement in clinical eye care practice to identify better and cheaper ways of identification, management, diagnosis and treatment of retinal disease. The importance of diabetic retinopathy screening programs and difficulty in achieving reliable early diagnosis of diabetic retinopathy at a reasonable cost needs attention to develop computer-aided diagnosis tool. Computer-aided disease diagnosis in retinal image analysis could ease mass screening of population with diabetes mellitus and help clinicians in utilizing their time more efficiently. The recent technological advances in computing power, communication systems, and machine learning techniques provide opportunities to the biomedical engineers and computer scientists to meet the requirements of clinical practice. Diverse and representative retinal image sets are essential for developing and testing digital screening programs and the automated algorithms at their core. To the best of our knowledge, the database for this challenge, IDRiD (Indian Diabetic Retinopathy Image Dataset), is the first database representative of an Indian population. Moreover, it is the only dataset constituting typical diabetic retinopathy lesions and also normal retinal structures annotated at a pixel level. This dataset provides information on the disease severity of diabetic retinopathy, and diabetic macular edema for each image. This makes it perfect for development and evaluation of image analysis algorithms for early detection of diabetic retinopathy.

Dataset Description: 1. Worldwide, eye ailments are recognized as significant contributors to nonfatal disabling conditions. In Bangladesh, 1.5% of adults suffer from blindness, while 21.6% experience low vision. Therefore, eye disease detection is crucial for preserving vision, preventing blindness, and maintaining overall health. Early detection allows for prompt intervention and treatment, preventing irreversible damage and preserving quality of life. By analyzing the dataset, researchers will be able to identify trends, develop algorithms for diagnosis, assess treatment effectiveness, and inform preventive measures.

1. Currently, computer vision methods show great promise in carrying out classification and detection tasks of this nature.

2. To develop computer vision-based algorithms, an extensive eye disease dataset is presented containing original and augmented datasets of a variety of eye diseases such as Retinitis Pigmentosa, Retinal Detachment, Pterygium, Myopia, Macular Scar, Glaucoma, Disc Edema, Diabetic Retinopathy, Central Serous Chorioretinopathy, and Healthy eye image. The classifications of this dataset are done with the help of a domain expert from a healthcare institute.

3. A total of 5335 images of healthy and affected eye images were collected from Anwara Hamida Eye Hospital in Faridpur and BNS Zahrul Haque Eye Hospital in Faridpur district with the help of the hospital authorities. Then from these original images, a total of 16242 augmented images are produced by using Rotation, Width shifting, Height shifting, Translation, Flipping, and Zooming techniques to increase the number of data.

Paper Link: A dataset of color fundus images for the detection and classification of eye diseases
Original Data Source: Eye Disease Image Dataset

Diabetic retinopathy (DR) is a late microvascular complication of Diabetes Mellitus (DM) that could lead to permanent blindness in patients, without early detection. Although adequate management of DM via regular eye examination can preserve vision in in 98% of the DR cases, DR screening and diagnoses based on clinical lesion features devised by expert clinicians; are costly, time-consuming and not sufficiently accurate. This raises the requirements for Artificial Intelligent (AI) systems which can accurately detect DR automatically and thus preventing DR before affecting vision. Hence, such systems can help clinician experts in certain cases and aid ophthalmologists in rapid diagnoses. To address such requirements, several approaches have been proposed in the literature that use Machine Learning (ML) and Deep Learning (DL) techniques to develop such systems. However, these approaches ignore the highly valuable clinical lesion features that could contribute significantly to the accurate detection of DR. Therefore, in this study we introduce a framework called DR-detector that employs the Extreme Gradient Boosting (XGBoost) ML model trained via the combination of the features extracted by the pretrained convolutional neural networks commonly known as transfer learning (TL) models and the clinical retinal lesion features for accurate detection of DR. The retinal lesion features are extracted via image segmentation technique using the UNET DL model and captures exudates (EXs), microaneurysms (MAs), and hemorrhages (HEMs) that are relevant lesions for DR detection. The feature combination approach implemented in DR-detector has been applied to two common TL models in the literature namely VGG-16 and ResNet-50. We trained the DR-detector model using a training dataset comprising of 1,840 color fundus images collected from e-ophtha, retinal lesions and APTOS 2019 Kaggle datasets of which 920 images are healthy. To validate the DR-detector model, we test the model on external dataset that consists of 81 healthy images collected from High-Resolution Fundus (HRF) dataset and MESSIDOR-2 datasets and 81 images with DR signs collected from Indian Diabetic Retinopathy Image Dataset (IDRID) dataset annotated for DR by expert. The experimental results show that the DR-detector model achieves a testing accuracy of 100% in detecting DR after training it with the combination of ResNet-50 and lesion features and 99.38% accuracy after training it with the combination of VGG-16 and lesion features. More importantly, the results also show a higher contribution of specific lesion features toward the performance of the DR-detector model. For instance, using only the hemorrhages feature to train the model, our model achieves an accuracy of 99.38 in detecting DR, which is higher than the accuracy when training the model with the combination of all lesion features (89%) and equal to the accuracy when training the model with the combination of all lesions and VGG-16 features together. This highlights the possibility of using only the clinical features, such as lesions that are clinically interpretable, to build the next generation of robust artificial intelligence (AI) systems with great clinical interpretability for DR detection. The code of the DR-detector framework is available on GitHub at https://github.com/Janga-Lab/DR-detector and can be readily employed for detecting DR from retinal image datasets.

Fundus (APTOS, DDR, IDRID, EYEPACS, Messidor) - Diabetic Retinopathy Classification

Dataset Overview This dataset is curated for diabetic retinopathy classification and combines images from five well-known fundus image datasets:

APTOS (Asia Pacific Tele-Ophthalmology Society) DDR (DeepDR) IDRID (Indian Diabetic Retinopathy Image Dataset) EYEPACS (A large-scale dataset used in DR detection challenges) Messidor (A benchmark dataset for DR detection) The dataset has been balanced to ensure an equal distribution of images across five severity classes.

Dataset Structure The dataset consists of three subsets:

Train: 5,968 images per class (Balanced) Validation: Separated for model tuning Test: Independent set for final evaluation Classes The dataset follows the standard diabetic retinopathy grading system, categorized into five classes:

No DR (0) – Healthy fundus images Mild DR (1) – Microaneurysms present Moderate DR (2) – More microaneurysms, hemorrhages, and exudates Severe DR (3) – Significant hemorrhages and vascular abnormalities Proliferative DR (4) – Neovascularization and severe damage Key Features Balanced Training Set: Each class has 5,968 images, preventing class imbalance issues. Multi-Source Data: Provides robust generalization by incorporating varied imaging conditions from multiple datasets. Real-World Relevance: Ideal for deep learning models in medical AI applications. Use Cases Training and evaluating deep learning models for automated DR classification. Transfer learning for medical image analysis. Benchmarking diabetic retinopathy detection models.

Over 1.5K images selected from the public Kaggle DR Detection dataset; Five DR grades (DR0 / DR1 / DR2 / DR3 / DR4), re-labeled by a panel of 45 experienced ophthalmologists; Eight retinal lesion classes, including microaneurysm, intraretinal hemorrhage, hard exudate, cotton-wool spot, vitreous hemorrhage, preretinal hemorrhage, neovascularization and fibrous proliferation; Over 34K expert-labeled pixel-level lesion segments; Multi-task, i.e., lesion segmentation, lesion classification, and DR grading.