Detection Of Retinal Diseases Research Articles

Multi-label Retinal Disease Classification Using Transformers.

Early detection of retinal diseases is one of the most important means of preventing partial or permanent blindness in patients. In this research, a novel multi-label classification system is proposed for the detection of multiple retinal diseases, using fundus images collected from a variety of sources. First, a new multi-label retinal disease dataset, the MuReD dataset, is constructed, using a number of publicly available datasets for fundus disease classification. Next, a sequence of post-processing steps is applied to ensure the quality of the image data and the range of diseases, present in the dataset. For the first time in fundus multi-label disease classification, a transformer-based model optimized through extensive experimentation is used for image analysis and decision making. Numerous experiments are performed to optimize the configuration of the proposed system. It is shown that the approach performs better than state-of-the-art works on the same task by 7.9% and 8.1% in terms of AUC score for disease detection and disease classification, respectively. The obtained results further support the potential applications of transformer-based architectures in the medical imaging field.

A Hybrid R-FTCNN based on principal component analysis for retinal disease detection from OCT images

Retinal disorders are of paramount importance in human life and if they do not receive treatment for retinal disorders, humans may undergo visual complications. The physician may diagnose more easily with optical coherence tomography (OCT) in ophthalmology, where early diagnosis is essential. To date, artificial intelligence has recently been implemented more and more frequently in the field of health thanks to technological advancements. In this study, a hybrid Retinal Fine Tuned Convolutional Neural Network (R-FTCNN) has been proposed for the detection of retinal diseases such as diabetic macular edema, drusen, and choroidal neovascularization from OCT images. In this study, the newly constructed R-FTCNN architecture and principal component analysis (PCA) are utilized concurrently as part of this methodology. PCA was used to convert the fully connected layers of the R-FTCNN to principal components, and the Softmax function was applied to the principal components to construct a new classification model. By creating various feature arrays from fully connected layers of R-FTCNN, ablation test was applied and FC layer (or combinations) with the best performance was determined. Afterwards, the FC layer with highest performance was used as the basis for the proposed methodology. Leveraging OCT datasets from Duke and California San Diego University (UCSD), which are often cited in the literature, the performance of the suggested technique was assessed. This technique had 100% accuracy, 100% sensitivity, 100% specificity, 100% precision, and 100% F1 score metric for classifying retinal disorders in the Duke dataset. Furthermore, this technique demonstrated 99.70% accuracy, 99.70% sensitivity, 99.90% specificity, 99.70% precision, and 99.70% F1 score metrics when classifying retinal disorders in the UCSD dataset. ROC curves were generated for the performance summary of the proposed method on two OCT datasets, and the area under the curve (AUC) metrics revealed that the suggested approach performed well in terms of classifications. In this study, pre-trained CNN architectures namely GoogLeNet, AlexNet, and ResNet (18, 50 and 101) were employed to compare the performance of the proposed technique. The proposed method outperformed state-of-the-art technology when conclusions from two OCT datasets were compared with those from past investigations. In conclusion, the proposed framework might aid in the creation of computer-aided diagnostic instruments for the early detection of retinal disorders in ophthalmology.

Domain Adaptation-Based Automated Detection of Retinal Diseases from Optical Coherence Tomography Images

Purpose To verify the effectiveness of domain adaptation in generalizing a deep learning-based anomaly detection model to unseen optical coherence tomography (OCT) images. Methods Two datasets (source and target, where labelled training data was only available for the source) captured by two different OCT facilities were collected to train the model. We defined the model containing a feature extractor and a classifier as Model One and trained it with only labeled source data. The proposed domain adaptation model was defined as Model Two, which has the same feature extractor and classifier as Model One but has an additional domain critic in the training phase. We trained the Model Two with both the source and target datasets; the feature extractor was trained to extract domain-invariant features while the domain critic learned to capture the domain discrepancy. Finally, a well-trained feature extractor was used to extract domain-invariant features and a classifier was used to detect images with retinal pathologies in the two domains. Results The target data consisted of 3,058 OCT B-scans captured from 163 participants. Model One achieved an area under the curve (AUC) of 0.912 [95% confidence interval (CI), 0.895–0.962], while Model Two achieved an overall AUC of 0.989 [95% CI, 0.982–0.993] for detecting pathological retinas from healthy samples. Moreover, Model Two achieved an average retinopathies detection accuracy of 94.52%. Heat maps showed that the algorithm focused on the area with pathological changes during processing, similar to manual grading in daily clinical work. Conclusions The proposed domain adaptation model showed a strong ability in reducing the domain distance between different OCT datasets.

AI telemedicine screening in ophthalmology: health economic considerations

AI telemedicine screening in ophthalmology: health economic considerations

Game Theory-Based Dynamic Weighted Ensemble for Retinal Disease Classification

An automated retinal disease detection system has long been in existence and it provides a safe, no-contact and cost-effective solution for detecting this disease. This paper presents a game theory-based dynamic weighted ensemble of a feature extraction-based machine learning model and a deep transfer learning model for automatic retinal disease detection. The feature extraction-based machine learning model uses Gaussian kernel-based fuzzy rough sets for reduction of features, and XGBoost classifier for the classification. The transfer learning model uses VGG16 or ResNet50 or Inception-ResNet-v2. A novel ensemble classifier based on the game theory approach is proposed for the fusion of the outputs of the transfer learning model and the XGBoost classifier model. The ensemble approach significantly improves the accuracy of retinal disease prediction and results in an excellent performance when compared to the individual deep learning and feature-based models.

Arterial hypertension and retinal layer thickness: the Beijing Eye Study

PurposeTo investigate relationships between blood pressure and the thickness of single retinal layers in the macula.MethodsParticipants of the population-based Beijing Eye Study, free of retinal or optic nerve disease, underwent...

LAC-GAN: Lesion attention conditional GAN for Ultra-widefield image synthesis

Automatic detection of retinal diseases based on deep learning technology and Ultra-widefield (UWF) images plays an important role in clinical practices in recent years. However, due to small lesions and limited data samples, it is not easy to train a detection-accurate model with strong generalization ability. In this paper, we propose a lesion attention conditional generative adversarial network (LAC-GAN) to synthesize retinal images with realistic lesion details to improve the training of the disease detection model. Specifically, the generator takes the vessel mask and class label as the conditional inputs, and processes the random Gaussian noise by a series of residual block to generate the synthetic images. To focus on pathological information, we propose a lesion feature attention mechanism based on random forest (RF) method, which constructs its reverse activation network to activate the lesion features. For discriminator, a weight-sharing multi-discriminator is designed to improve the performance of model by affine transformations. Experimental results on multi-center UWF image datasets demonstrate that the proposed method can generate retinal images with reasonable details, which helps to enhance the performance of the disease detection model.

A lightweight deep learning model for retinal optical coherence tomography image classification

AbstractAutomated retinal disease detection and grading is one of the most researched areas in medical image analysis. In recent years, Deep Learning models have attracted much attention in this field. Hence, in this paper, we present a Deep Learning‐based, lightweight, fully automated end‐to‐end diagnostic system for the detection of the two major retinal diseases, namely diabetic macular oedema (DME) and drusen macular degeneration (DMD). Early detection of these diseases is important to prevent vision impairment. Optical coherence tomography (OCT) is the main imaging technique for detecting these diseases. The model proposed in this work is based on residual blocks and channel attention modules. The performance of the model is evaluated using the publicly available Mendeley OCT dataset and the Duke dataset. We were able to achieve a classification accuracy of 99.5% in the Mendeley test dataset and 94.9% in the Duke dataset with the proposed model. For the application, we performed an extensive evaluation of pre‐trained models (LeNet, AlexNet, VGG‐16, ResNet50 and SE‐ResNet). The proposed model has a much smaller number of trainable parameters and shows superior performance compared to existing methods.

Retinal Vessel Segmentation, a Review of Classic and Deep Methods.

Retinal illnesses such as diabetic retinopathy (DR) are the main causes of vision loss. In the early recognition of eye diseases, the segmentation of blood vessels in retina images plays an important role. Different symptoms of ocular diseases can be identified by the geometric features of ocular arteries. However, due to the complex construction of the blood vessels and their different thicknesses, segmenting the retina image is a challenging task. There are a number of algorithms that helped the detection of retinal diseases. This paper presents an overview of papers from 2016 to 2022 that discuss machine learning and deep learning methods for automatic vessel segmentation. The methods are divided into two groups: Deep learning-based, and classic methods. Algorithms, classifiers, pre-processing and specific techniques of each group is described, comprehensively. The performances of recent works are compared based on their achieved accuracy in different datasets in inclusive tables. A survey of most popular datasets like DRIVE, STARE, HRF and CHASE_DB1 is also given in this paper. Finally, a list of findings from this review is presented in the conclusion section.

Weakly-Supervised Learning With Complementary Heatmap for Retinal Disease Detection.

There are many types of retinal disease, and accurately detecting these diseases is crucial for proper diagnosis. Convolutional neural networks (CNNs) typically perform well on detection tasks, and the attention module of CNNs can generate heatmaps as visual explanations of the model. However, the generated heatmap can only detect the most discriminative part, which is problematic because many object regions may exist in the region beside the heatmap in an area known as a complementary heatmap. In this study, we developed a method specifically designed multi-retinal diseases detection from fundus images with the complementary heatmap. The proposed CAM-based method is designed for 2D color images of the retina, rather than MRI images or other forms of data. Moreover, unlike other visual images for disease detection, fundus images of multiple retinal diseases have features such as distinguishable lesion region boundaries, overlapped lesion regions between diseases, and specific pathological structures (e.g. scattered blood spots) that lead to mis-classifications. Based on these considerations, we designed two new loss functions, attention-explore loss and attention-refine loss, to generate accurate heatmaps. We select both "bad" and "good" heatmaps based on the prediction score of ground truth and train them with the two loss functions. When the detection accuracy increases, the classification performance of the model is also improved. Experiments on a dataset consisting of five diseases showed that our approach improved both the detection accuracy and the classification accuracy, and the improved heatmaps were closer to the lesion regions than those of current state-of-the-art methods.

Prediction of Retinal diseases using Convolutional Neural Networks

Abstract: Eye diseases are the common diseases people are facing nowadays. Even people of all ages ranging from children to old are affected due to eye diseases. Some diseases will become very complex if they are untreated at an early stage. To overcome such problems early detection and treatment is required which could prevent vision loss. For many, eye disorders have become almost a part of their daily lives. Traditionally, retinal disease detection techniques were based on manual description. The WHO (World Health Organization) finds that there are millions of visually impaired people all over the world. The retinal diseases include Cataract, Crossed Eyes, Bulging Eyes, Uveitis and Glaucoma can be predicted using convolutional neural networks in this project. Medical Imaging has a huge scope for the application of recent advancements in computation. With the emerging computer technology, the development of an automatic system for diseases using Convolutional neural Networks in association with different optimization methods (SGDM, RMS-PROP, and ADAM). Treating the disease is now possible especially if the patient’s area and medical services are limited. Thus, the idea is to build an algorithm that automatically identifies whether a patient is suffering from any eye diseases or not by looking at the images

Detection of retinal diseases from ophthalmological images based on convolutional neural network architecture

The retina is an eye layer that incorporates light- and color-sensitive cells as well as nerve fibers. It collects light and distributes it to the brain for image processing through the use of the optic nerve. Diseases that end up causing vision loss and blindness are generated by retinal ailments. As a result, it is imperative to diagnose and treat certain disorders as early as possible. Optical coherence tomography (OCT), an angiography imaging technique, is operated to help diagnose retinal disorders. Deep learning approaches, which are extensively utilized, have now become a convenient way for diagnosing retinal illnesses through OCT images as a result of their effective outcomes in interpreting medical images. To diagnose retinal disorders utilizing OCT scans, this investigation developed a hybrid methodology based on image pre-processing and convolutional neural networks (CNNs) (a deep learning method). Image pre-processing techniques including background filling, resizing, noise reduction, and highlighting are exercised at the pre-processing stage. The segmentation process provides a new CNN architecture with five convolution layers that does have a low computational cost. Compared to other publications using the same data set, the proposed method seems to have a success rate of 99.48 percent in the detection of retinal disorders, closing a significant gap in the literature. The proposed approach has the advantage of maintaining low computing costs in comparison to other studies in the literature. When the conclusions are regarded, it is noticed that the suggested method might be exerted as a decision support system to assist physicians in the clinical context during the diagnosis of retinal disorders.

Digital health in medicine: Important considerations in evaluating health economic analysis

Digital health in medicine: Important considerations in evaluating health economic analysis

Use of dominant activations obtained by processing OCT images with the CNNs and slime mold method in retinal disease detection

Retinal disease is one of the diseases that cause visual symptoms or loss of vision in humans. This disease can affect the choroid, which severely affects vision. Optical coherence tomography (OCT) images are usually used to detect retinal disease. OCT is an imaging technique that takes high-resolution slices of retinal images. It takes time for experts to examine and interpret the OCT images. Experts need to take advantage of technological capabilities to make this process faster and more accurate. Three datasets were used in this study. Dataset #1 (UCSD dataset) consists of choroidal neovascularization (CNV), diabetic macular edema (DME), drusen, and normal OCT image types. Dataset #2 (Duke dataset) and Dataset #3 consist of age-related macular degeneration (AMD), DME, and normal OCT image types. An artificial intelligence based hybrid approach was proposed for retinal disease detection. In the proposed approach, class-based activations were extracted for each model with nine transfer learning models using the dataset. Next, the dominant activations were selected from the model-based activations of each class using the slime mold algorithm (SMA) and the selected activations were classified using the softmax method. The overall accuracy obtained in classification is as follows: 99.60% for dataset 1, 99.89% for dataset #2 and 97.49% for dataset #3. In this study, it was found that the proposed approach contributes to the performance of transfer learning models.

Detecting retinal vasculature as a key biomarker for deep Learning-based intelligent screening and analysis of diabetic and hypertensive retinopathy

Retinal vessels are considered important biomarkers for the detection of retinal diseases, like diabetic retinopathy (caused by diabetes) and hypertensive retinopathy (caused by hypertension). The manual finding from this retinal vasculature is time-consuming and costly. The image quality of the fundus image directly affects the accurate segmentation of these vessels in the automatic methods. With such inferior quality images, deep-learning-based methods are better for dealing with segmentation. Conventional deep-learning-based vessel segmentation methods deal the segmentation task with deeper convolutional neural networks and many trainable parameters. Minor changes in the retinal vasculature, such as those that result from the creation of new smaller vessels, is crucial for the keen analysis of diseases (e.g., diabetic retinopathy). The small vessels are crucial to segment, owing to continuous max-pooling operations and deeper networks. We herein present a pool-less residual segmentation network that is capable of segmenting even smaller vessels using a shallower network with a low number of trainable parameters. Our proposed pool-less residual segmentation network (PLRS-Net) is a vessel segmentation network that provides the pooling effect with strided convolution for better segmentation sensitivity. The final PLRS-Net is an advanced form of a pool-less segmentation network (PLS-Net) wherein semantic segmentation was performed with a few layers, and the residual connection fulfilled the feature enhancement strategy to construct PLRS-Net. PLS-Net and PLRS-Net are two separate networks that can perform vessel segmentation without prior preprocessing and postprocessing.To evaluate our proposed method, the experiments include three publicly available datasets: Digital retinal images for vessel extraction (DRIVE), child heart health study in England database (CHASE-DB1), and structured analysis of retina (STARE). The results demonstrate that our proposed method provides a high segmentation performance, achieving an average sensitivity (Sen) of 82.69, specificity (Spe) of 98.17, accuracy (Acc) 96.82, and area under the curve (AUC) of 98.35 for the DRIVE dataset, Sen of 83.01, Spe of 98.39, Acc of 97.31, and AUC of 98.63 for the CHASE-DB1 dataset, and a Sen of 86.35, Spe of 98.03, Acc of 97.15, and AUC of 98.99 for the STARE dataset. These accuracies show exceptional segmentation performance of the proposed method compared to state-of-the-art approaches for automatic vessel detection for diagnosis purposes.

A multi-scale anomaly detection framework for retinal OCT images based on the Bayesian neural network

As a non-contact imaging technology, optical coherence tomography (OCT) is widely used in retinal disease detection. There is a growing interest among researchers for automated detection of retinal diseases using OCT images. With the development of deep learning, convolutional neural network (CNN) enables accurate detection of common retinal diseases. However, large-scale labeled data is necessary for these deep learning methods. It limits the application of these methods in retinal disease detection, especially for diseases with low occurrence. Therefore, we propose an automated anomaly detection method for unspecific retinal disease, and only healthy OCT images are needed for training. The method uses epistemic uncertainty as the criterion of anomaly detection. Our method assumes that the lesion correlates with a high epistemic uncertainty region. We propose a Bayesian neural network (BNN) model, Multi-scale Bayesian U-Net (MBU-Net), to obtain epistemic uncertainty of OCT images. We then design an algorithm to reduce the uncertainty generated by healthy tissue regions. Finally, a threshold-based function is designed to distinguish whether the input data is healthy or not. We evaluate our approach on a public dataset, UCSD dataset, and a private dataset collected by ourselves, named BFHJLU dataset. The experimental results show that the proposed method can achieve 94.9% accuracy, 97.9% sensitivity, and 86.0% specificity on the UCSD dataset. On the BFHJLU dataset, the accuracy, sensitivity, and specificity are 92.1%, 97.3%, and 87.7%. The proposed method outperforms existing anomaly detection methods.

Retinal damage alters gene expression profile in lacrimal glands of mice

Early detection of such retinal diseases as glaucoma and age-related macular degeneration (AMD) is important to prevent blindness. There have been reports of changes in some components in the tears of glaucoma and AMD patients, suggesting tears' potential usefulness in screening for retinal diseases. We hypothesized that retinal damage might alter gene expression in the lacrimal gland, leading to those changes in tear components. We caused retinal damage in mice by intravitreal injection of N-methyl-d-aspartate (NMDA) or excessive light exposure. Hematoxylin and eosin staining showed no histological changes in the lacrimal glands of animals whose retinas had been damaged. However, RNA sequencing of lacrimal glands on the 3rd day after NMDA injection or light exposure revealed changes in the expression of 491 genes (268 up-regulated; 223 down-regulated) in the NMDA group and 531 genes (311 up-regulated; 220 down-regulated) in the light group. Further gene-set enrichment analysis indicated that both types of retinal damage activated the immune system in the lacrimal glands. This is the first demonstration that retinal damage can alter gene expression in the lacrimal glands, and it might lead to a novel non-invasive screening method for early detection of retinal diseases.

A ROBUST TECHNIQUES OF ENHANCEMENT AND SEGMENTATION BLOOD VESSELS IN RETINAL IMAGE USING DEEP LEARNING

The retina is the most important part of the eye. Early detection of retinal disease can be done through the passage of the blood vessels of the retina. Enhancement of the quality of retinal images that have both noise and noise is the first step in image processing to help improve the accuracy of the results for image segmentation and extraction. Images store a lot of information, but often there is a decrease in quality or image defects. So that images that have experienced interference or noise are easily interpreted, then the image can be manipulated into other images of better quality using image processing techniques or methods. The neural network-based method that is currently popular is deep learning. The segmentation process is currently a widely used method of deep learning that has grown rapidly used in various studies. One of the popular methods is Convolutional Neural Network (CNN). CNN can handle large-dimensional data such as images because the input to CNN is in the form of a matrix. Since the findings of retinal blood vessel segmentation are often inaccurate and there is always noise, this study will look at how to segment retinal images in blood vessels using CNN U-Net and LadderNet methods. Proper segmentation of retinal blood vessels can be the first step to detecting a disease. Segmentation and analysis of retinal blood vessels can assist medical personnel in detecting the severity of a disease. The stages of image enhancement used are Histogram Equalization and Clahe. Segmentation of blood vessels is done using CNN U-Net and LadderNet Methods. The results of the application of the enhancement and segmentation using the U-Net and LadderNet methods on training and on testing data were tested on the DRIVE dataset. The results of measurement of accuracy, specificity, sensitivity and F1 Score of blood vessel segmentation using the U-Net CNN method were 95.46%, 98.56%, 74.20%, and 80.63%, respectively. While the results of the CNN LadderNet method were 95.47%, 98.42%, 75.19%, and 80.86%, respectively. Based on the results of blood vessel segmentation from two proposed methods, the result of the CNN LaddetNet method is greater than the CNN U-Net method in accuracy, sensitivity, and F1 Score. The proposed approach will be further developed in the future, with the aim of increasing the value of the blood vessel segmentation process evaluation outcomes.

Coherent convolution neural network based retinal disease detection using optical coherence tomographic images

An optical coherence tomography images are used to visualize the retinal micro-architecture and perform an easy scan of its abnormalities. In this paper, a coherent convolutional neural network is proposed for four-class classification of retinal diseases and able to detect neovascularization (CNV), diabetic macular edema (DME), DRUSEN, and NORMAL class label in the OCT images. The new proposal overcomes three of the challenges by (1) more profoundly detect the irregular patterns of each class of retinal disease (2) manages consistency between input and output of the network (3) cohesively bound the layers of the network for easy flow of image features. The proposed convolution neural network model is having five layers. In order to adopt coherent behavior, the proposed model inculcating the batch normalization layer along with the every activity layer and obtained an accuracy of 97.19% for retinal disease detection. Moreover, the performance of this method is remarkably good as compared to other standard deep learning methods. This proposal is a promising step in revolutionizing the present scenario of ocular diagnostic system and has the potential to generate a significant clinical impact.

Automatic Detection of Retinal Diseases Based on Lightweight Convolutional Neural Network

Automatic Detection of Retinal Diseases Based on Lightweight Convolutional Neural Network