Search

Diabetes mellitus, also known as diabetes, is a metabolic disease which may cause high blood sugar, blurred vision and several other complications. Patients already suffering from diabetes have a greater probability to develop Diabetic Retinopathy. Most prominently in the blood vessels of the tissue residing in the rear portion of the eye gets damaged causing vision problems. This disease is especially prevalent in India. Test for Diabetic Retinopathy involves a dilated eye exam, which is time consuming and requires specialized doctors and distinct equipment which may not be easily accessible to the people especially in rural areas. Predicting the possibility of Diabetic Retinopathy utilizing fundus image is considered an effectual analysis subject by many. In this paper a method is explored to diagnose Diabetic Retinopathy using fundus images and deep learning to accurately tell the patient whether he has a Diabetic Retinopathy. Features used from the eye fundus images are soft exudate and hard exudate along with optic disc. A website is also made to deploy the model and promote mass diagnosis at a faster pace.

Unprecedented modern research in medical science reveals a new horizon in detection, diagnosis, and treatment of the various ophthalmology-related disorders by analyzing vessel configurations in retinal fundus images. The main motivation is to come up with a new scheme to separate out vessels from the others in fundus image without manual intervention. In the suggested method, first, we generate three layers from the fundus with different features namely ‘prominent vessel network’, ‘minute vessel details’, and ‘illumination improved fundus’. In the second step, a new algorithm has been defined to clean unwanted ‘dust-like’ noise and restructure the vascular network with all the subtle branches by means of a binary image. The new scheme is executed and validated on images available in open-sourced DRIVE repository. Outcomes of the proposed scheme are tallied with other schemes for establishing its superiority.

Diagnosing and detection of ophthalmology related disorders, extraction of vessels from retinal fundus images is very important task in Medical Science. Locating retinal vessels manually is not only tough and time taking task for ophthalmologist but also it is prone to have human error because of complex nature of blood vessels in Retina. Across the globe different techniques are being applied to overcome this challenge by deploying different strategies and techniques. This paper proposes a new method for extraction of vessels from retinal images. Proposed method first extracts thick vessels from the fundus image. Finally, flood filling algorithm in-conjugation with BFS is applied to discover the tiny blood vessels and reconstruct the final result by combining it with the thick vessels extracted in the first place. Experiment is performed on publicly available DRIVE data set and it is found that proposed method produced quite outstanding results.

Analysis of images with deep learning and machine learning has recently show more accuracy in medical image processing. The paper review the major machine learning methods is used for the extraction of retinal fundus image. It describes how hardware has been modified according to time for retinal eye segmentation that device is helpful for the environment and also useful for humans The algorithms used for image extraction, clear and noise-free Images of different color scales are used to improve image quality. The paper gives a review of how researchers and scientists have been trying different methods to improve vessel segmentation in the future.

At least one person in every 20 people is a diabetic patient all over the world. Long term diabetes may cause diabetic retinopathy (DR), which leads to complete blindness except it is identified at the initial stage. Hard exudates are one of the most noticeable symptoms for DR. In this paper, an automated process to spot the presence of hard exudates in a retinal fundus image has been developed. This work will be helpful to ophthalmologists to detect DR in a rapid and reliable way even at the early stage. This proposed methodology is a generic process as it does not require any training dataset to recognize the exudates. The rudiments of digital image processing have been utilized here to locate the hard exudates. Telemedicine has become an indispensable thing after the expansion of COVID-19. Reliability on automated reports is very important to provide a proper service through telemedicine. A dual watermarking process has been employed over the automated test image by using an imperceptible and fragile mark to provide integrity, and a robust visible mark for content labeling. The efficiency of the whole process has been optimized in all possible aspects.

Cardiovascular disease is now the most common cause of death worldwide. According to the Centre’s for Disease Control (CDC), heart disease is the most leading cause of the death in the United States, United Kingdom, Australia, and Canada. Nearly half of all U.S. adults have some type of cardiovascular disease. A common microvascular complication, diabetic retinopathy (DR), has been repeatedly shown to be associated with an elevated risk of cardiovascular disease (CVD). In India, the diabetic population in the country is close to hitting the alarming mark of 69.9 million by 2025 and 80 million by 2030. Medical discoveries are made by observing relationships and making hypotheses from those observations then after designing them and then running the experiments to test the hypotheses generated. Here, we show that the deep learning can extract new knowledge from the retinal fundus images of the diabetic retinopathy patients. We use deep learning models on the patient’s data collected from various hospitals with the symptoms of diabetic retinopathy. This paper predicts the diabetic retinopathy from retinal fundus images by using deep learning convolutional neural network. This paper yields an accuracy of 89 percent in the identification of diabetic retinopathy (DR) from the given retinal fundus image of the patient.

Diabetic retinopathy (DR) is a microvascular symptom where retina is affected by fluid leaks of the fragile blood vessels. Clinically, retinal Hemorrhages are one of the earliest indications of diabetic retinopathy disease. In this contrast, the Hemorrhage count is used to indicate the severity of this disease. The early detection of retinal Hemorrhages obviously prevents the incurable blindness of the DR patients. But, retinal Hemorrhage detection is still a challenging task. Highly reliable, accurate, platform independent retinal Hemorrhage detection method is still an open field. In this research article, we have reviewed the principal methodologies which are used to diagnose the retinal Hemorrhages under the diabetic retinopathy screening operations. This review article helps the researchers to develop a high quality retinal Hemorrhage screening method in future.

The detection of microaneurysms at an early stage plays an imperative role to reduce the severity of diabetic retinopathy (DR) disease. As microaneurysm is a common sign of diabetic retinopathy, which appears as a red lesion in the fundus image. This paper deals with the classification of microaneurysms and non-microaneurysms. The proposed work follows four steps, specifically pre-processing, candidate extraction, extraction of texture features from candidate region using Gabor filter and finally classification based on extracted features. Support vector machine (SVM) has been used for classification of microaneurysms (MAs). The proposed approach has been tested on Indian diabetic retinopathy image dataset (IDRiD). The reported result of accuracy is 80.85%, specificity is 78.76% and sensitivity 76.75% highlights the effectiveness of the presented technique.

Diabetes retinopathy is a consequence of diabetes mellitus. It is one of the major reasons of avoidable blindness in rural as well as urban areas. This disease is mainly caused if body is not able to produce enough insulin, causing high blood sugar levels. The minute vessels in the retinal areas get swollen due to high sugar levels in blood, resulting them to rapture and spread of blood in the retinal region. This results in to haemorrhages, i.e. red lesions in retina. With the advancements in the field of image processing, these lesions are easily detected at earlier stage to avoid further consequences. In this research work, the morphological approach has been developed to detect these lesions. This mainly includes pre-processing, detection and elimination of fovea and blood vessels. The microaneurysms being minute lesions are eliminated to detect haemorrhages. The proposed work is carried out on Indian diabetic retinopathy image dataset (IDRiD). The accuracy, sensitivity and specificity achieved with this approach is noted as 98.22%, 75.3% and 91.65%, respectively.

Diabetic retinopathy has emerged as one of the prime reasons for loss of vision. It is diagnosed by analyzing the abnormalities in retinal fundus image. In diabetic patients, the blood vessels become abnormal over time, which results in blockages. These abnormalities lead to the development of various types of aberrations in the retina. High sugar levels make the blood vessels defective and result in formation of bright and dark lesions. The risk of loss of vision is reduced by detecting lesions by analyzing the fundus image in the early stage of diabetic retinopathy. This chapter reviews various studies for automatic abnormality detection in fundus images with a purpose of easing out the work of researchers in the field of diabetic retinopathy. A condensed study on methodology and performance analysis of each detection algorithm is laid out in a simple table form, which can be reviewed effortlessly by any researcher to realize the advantages and shortcomings of each of these algorithms. This review highlights various research protocols for detection and classification of retinal lesions. It provides guidance to researchers working on retinal fundus image processing.

Supervised machine learning algorithm based retinal hemorrhage detection and classification is presented. For developing an automated diabetic retinopathy screening system, efficient detection of retinal hemorrhage is important. Splat, which is a high level entity in image segmentation is used to mark out hemorrhage in the pre-processed fundus image. Here, color images of retina are portioned into different segments (splats) covering the whole image. With the help of splat level and GLCM features extracted from the splats, two classifiers are trained and tested using the relevant features. The ground-truth is established with the help of a retinal expert and using dataset and clinical images the validation was done. The trained classifier’s output is evaluated and the classifier with the best output is chosen for implementation in hardware.

There are countless number of people with diabetes around the world. diabetic retinopathy (DR), a major complication of diabetes, is a retinal disease that results in blindness if not treated in the earlier stages. Early detection of diabetic retinopathy in patients can help avoid further loss of eyesight. For early detection, it is important for diabetic patients to undergo medical tests after a frequent amount of times. Diagnosing DR using fundus images is a difficult and time-consuming task which requires professional expertise to find and identify if the DR traits are present in the eye. In our proposed system, we make use of artificial intelligence, convolutional neural networks precisely to make the process of detection and classification of diabetic retinopathy a lot simpler and fast. The proposed model makes use of AlexNet which is a convolutional neural network architecture which is trained based on fundus image database to accurately diagnose DR with minimum efforts. The database contains multiple cases of retinal hemorrhage, microaneurism, cotton wool spots, etc. The retinal image of the patient is fed to the system where initially the image is trimmed to match the input layer of AlexNet and then the neural network extracts features and the model gets trained in order to further classify the diabetic retinopathy patients. The last layer of the convolutional neural network is developed in such a way each segment of image is assigned a specific weight factor where higher the weight, more is the chances of getting severe DR.

Peri-Papillary Atrophy (PPA) is a clinical finding related to chorioretinal diminishing and disorder analysis of the Retinal Pigment Epithelium (RPE) in the region encircling the optic disk. Optical Coherence Tomography (OCT) is an integral asset in clinical perform, used to survey retinal harm in a scope of illnesses. It gives ultra-high-resolution cross-sectional pictures of organic tissues utilizing nonobtrusive imaging innovation. Be that as it may, the examination by OCT is still less open in view of the astonishing costs in various ophthalmology centers wherever all through the world. The primary prerequisite for Peri-papillary-decay investigation in glaucoma analysis is early recognition of RNFL deterioration so as to arrange the treatment process as quickly as time permits. Henceforth, a mass screening agenda is by all accounts appropriate for supporting the finding. Fundus imaging is inexpensive and easily accessible rather than OCT imaging. The contemporary medical accomplishment needs to focus on the analysis procedure basis on the fundus image for the fulfillment of the huge demand of screening process for Glaucoma. In our work we are focusing on Peri-Papillary Atrophy Analysis (PPAA) using Retina Nerve Fiber Layer (RNFL) analysis using Fundus image of retina. In RNFL analysis, we need to identify RNFL loss. This RNFL misfortune can be moderately all around demonstrated as a surface change in fundus photos by analyzing the heterogeneous idea of the Peri-Papillary Atrophy area. RBF kernel based SVM classifier using LDA in combination is applied to achieve the motive of this literature. The experimental results reveal the benefits of the classifier worked on fourteen statistical features.

Deep learning performs as a computational tool with various potential utilities in ophthalmology. Retinal infections of the eye need to analyze small retinal vessels, microaneurysms, and exudates in the diagnosis of retinal diseases. Due to the appearance of various noises in the fundus images, the retinal vasculature is too complicated to be analyzed for retinal conditions. In this work, we have focused on the field of advanced deep learning in which plethora of architecture is available with the increase in dimension and flexibility of the retinal fundus images. Removal of noise is an essential part to better visibility of noisy fundus and thus a deep learning method for degraded retinal fundus image restoration scheme has been suggested in this investigation. A deep convolutional denoising auto-encoder method based on total variational multi-norm loss function minimization with batch normalization approach has been introduced for restoration of the fundus. The proposed scheme is utilized to restore the perceptible structural details of fundus as well as to decrease the noise level. Moreover, the speed of the network for target noisy images is faster compared to that of other models after fine-tuning of the network with the dropout mechanism. The retinal image databases such as DRIVE, STARE, and DIARETDB1 have been adopted to assess image denoising effects. Our approach to increase the visibility of fundus images by noise reduction through a deep training method has significantly delivered better performance without losing image details along with having fast convergence rate.

Retinal vascular angles are found to be one of the crucial parameters for biometric template generation. Many of the previous literature have used bifurcation angles for arteriovenous classification also. Acute angles are preferred as bifurcation angles in most of the cases. To prove vascular angle at a bifurcation point as an acute angle is itself a challenging task. In this paper, an initiative has been taken to prove bifurcation angle is acute followed by distinguishing the actual and distorted bifurcation angle. The accuracy of the proposed approach has been evaluated from a publicly available database using the experimental observation.

Present days, the development of computer-aided diagnosis of diabetic retinopathy (DR) is an active research in the field of medical image processing. The true quality performance of medical image understanding classifier depends on tested fundus database by some means. Therefore, a researcher must have the knowledge of all types of database available to facilitate the development of DR screening tool and this paper is intended to provide an idea of various fundus databases. Trends are seen that a number of researchers use different databases to evaluate their code, in that way the purpose to attain universally accepted diabetic retinopathy screening tool will not be achieved at earliest. However, there are no unique databases available to evaluate the program for all types of DR symptoms on fundus image. In this paper, present the detail comparative discussion on various publicly available databases of fundus images, which established to facilitate to do research work for the development of computer-aided diabetic retinopathy screening tool (CAD-DRST). The selection of test database or multiple databases depends on the researcher and the quality of the database; along with test results signify the acceptability level of the proposed algorithms. Besides that, through this paper, we urged to all scientific communities to establish a common database for the finest development of CAD-DRST.

Almost 6.5 million people are affected due to glaucoma, diabetics every year in the USA (CDC, Diabetic retinopathy. National for chronic disease prevention and health promotion [Online]. Atlanta, GA [1]). The paper presents a computer-aided blood vessel detection from fundus imaging. This is mainly necessary for premature baby and diabetic patient (Karperien et al. Clin Ophthalmol 2(1):109 [2]). By proper screening of the vessel of patients, the medicine can be modified accordingly. For this case, DRIVE database (Staal et al. IEEE Trans Med Imaging 23(4):501–509 [3]) is chosen. And this detection scheme is fully automated and initially based on top-hat filtering on the input image. Iteratively, region growing and median filtering are applied on the initial segmented output and result in 94.02% accuracy when compared with the ground truth. Diabetic retinopathy analysis and retinal vessel tortuosity in premature infants can be determined using the proposed approach in this paper.

Cloud-Assisted Electronic Eye Doctor for automatic glaucoma screening which uses an expert system is proposed. The expert system is used for diagnosis, which uses databases of specialist knowledge to make decisions or offer advice to user or doctor. Data including patient’s medical state and fundus image are generated, medical records are gathered and delivered through the cloud tier in the cloud, and the investigation of fundus images is performed. A cost-effective mean for glaucoma screening is presented facilitated by all over access nature of the system via cloud platform allowing the disease to be discovered earlier and enabling early examination for more efficient intervention and disease management.

A fundus image plays a significant role to analyze a wide variety of ophthalmic conditions. One of the major challenges faced by ophthalmologist in the analysis of fundus images is its low contrast nature. In this paper, two stage histogram enhancement schemes to improve the visual quality of fundus images are proposed. Fuzzy logic and Histogram Based Enhancement algorithm (FHBE) and Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm are cascaded one after the other to accomplish the two stage enhancement task. This results in two new enhancement schemes, namely FHBE-CLAHE and CLAHE-FHBE. The analysis of the results based on its visual quality shows that two stage enhancement schemes outperforms individual enhancement schemes.

Analysis of the retinal blood vessels has become remarkable area of research in biomedical field. This paper presents fundus image blood vessel segmentation approach using region-based features. In the pre-processing phase, the input fundus image is segmented as major vessel and minor vessel region. Further, to enhance segmentation accuracy region-based features are extracted from minor vessels by applying morphological operations. Fuzzy entropy measure is used to select the relevant features and for classification, a k-NN classifier is employed. The proposed algorithm is evaluated using two openly available data sets DRIVE and CHASE_DB1. The method presented is independent of training samples and achieves 96.75% of classification accuracy.

With the immense acceptance and adoption of personal identification using biometrics, exigency of a reliable, faster and less expensive authentication process has arrived. In this novel work, we have attempted to exhibit some crucial features of human retina which are sufficient to build a secured biometric template in considerable amount of time, ignoring the other hazardous factors which may lead to even authentication failure. Experimental results prove that this proposed work would be able to identify the distinct bifurcation points in retinal fundus image, avoiding the crossovers consciously as they are not anatomically stable and could change their locations as an effect of some diseases. This work will reduce the complexity of any authentication algorithm drastically by concentrating only on existing bifurcations, ignoring tedious calculations on every junction points.

The unexpected changes in the width of retinal blood vessels can be recognized as one of the significant characters during the proliferation of several diseases. Clinical studies verify that width of retinal blood vessels may undergo changes during diabetes, hypertension, Retinopathy of prematurity and Proliferative Diabetic Retinopathy. Optical vessel swelling or Papilledema and optical neuritis are prominent ocular diseases that contribute to changes in the width. In this approach, width of retinal blood vessels has been determined to benefit the identification of such diseases. Every retinal fundus images exhibits several unique bifurcation points. The bifurcation points ensure stability and are not subject to any change even when prone to diseases. Identifying these points, our algorithm attempts to find the ratio of the width of a parent vessel with its corresponding child vessels at the bifurcation points. Experimental results prove that this ratio lies in the range of 1.1–1.8. Abrupt changes in the ratio ensures the probability of existence of ocular diseases.

Reliable detection of fundus lesion is important for automated screening of diabetic retinopathy. This paper presents a novel method to detect the fundus lesion in retinal fundus image based on a visual attention model. The proposed method intends to model the visual attention mechanism of ophthalmologists during observing fundus images. That is, the abnormal structures, such as the dark and bright lesions in the image, usually attract the most attention of experts, however, the normal structures, such as optic disc and vessels, have been usually selectively ignored. To measure the visual attention for abnormal and normal areas, the incremental coding length is computed in local and global manner respectively. The final saliency map of fundus lesion is a fusion of attention maps computed for the abnormal and normal areas. Experimental results conducted on the publicly DiaRetDB1 dataset show that the proposed method achieved a sensitivity of 0.71 at a specificity of 0.82 and an AUC of 0.76 for fundus lesion detection, and achieved an accuracy of 100 % for normal area (optic disc) detection. The proposed method can assist the ophthalmologists in the inspection of fundus lesion.