L-Seg: An end-to-end unified framework for multi-lesion segmentation of fundus images

Prior-guided attention fusion transformer for multi-lesion segmentation of diabetic retinopathy

To solve the issue of diagnosis accuracy of diabetic retinopathy (DR) and reduce the workload of ophthalmologists, in this paper we propose a prior-guided attention fusion Transformer for multi-lesion segmentation of DR. An attention fusion module is proposed to improve the key generator to integrate self-attention and cross-attention and reduce the introduction of noise. The self-attention focuses on lesions themselves, capturing the correlation of lesions at a global scale, while the cross-attention, using pre-trained vessel masks as prior knowledge, utilizes the correlation between lesions and vessels to reduce the ambiguity of lesion detection caused by complex fundus structures. A shift block is introduced to expand association areas between lesions and vessels further and to enhance the sensitivity of the model to small-scale structures. To dynamically adjust the model's perception of features at different scales, we propose the scale-adaptive attention to adaptively learn fusion weights of feature maps at different scales in the decoder, capturing features and details more effectively. The experimental results on two public datasets (DDR and IDRiD) demonstrate that our model outperforms other state-of-the-art models for multi-lesion segmentation.

A Classification-Based Adaptive Segmentation Pipeline: Feasibility Study Using Polycystic Liver Disease and Metastases from Colorectal Cancer CT Images

Automated segmentation tools often encounter accuracy and adaptability issues when applied to images of different pathology. The purpose of this study is to explore the feasibility of building a workflow to efficiently route images to specifically trained segmentation models. By implementing a deep learning classifier to automatically classify the images and route them to appropriate segmentation models, we hope that our workflow can segment the images with different pathology accurately. The data we used in this study are 350 CT images from patients affected by polycystic liver disease and 350 CT images from patients presenting with liver metastases from colorectal cancer. All images had the liver manually segmented by trained imaging analysts. Our proposed adaptive segmentation workflow achieved a statistically significant improvement for the task of total liver segmentation compared to the generic single-segmentation model (non-parametric Wilcoxon signed rank test, n = 100, p-value << 0.001). This approach is applicable in a wide range of scenarios and should prove useful in clinical implementations of segmentation pipelines.

Lightweight and multi-lesion segmentation model for diabetic retinopathy based on the fusion of mixed attention and ghost feature mapping

Diabetic retinopathy is the main cause of blindness, and lesion segmentation is an important basic work for the diagnosis of this disease. The main lesions include soft and hard exudates, microaneurysms, and hemorrhages. However, the segmentation of these four types of lesions is difficult because of their variability in size and contrast, and high intertype similarity. Currently, many network models have problems, such as a large number of parameters and complex calculations, and most segmentation models for diabetic retinopathy focus only on one type of lesion. In this study, a lightweight algorithm based on BiSeNet V2 was proposed for the segmentation of multiple lesions in diabetic retinopathy fundus. First, a hybrid attention module was embedded in the semantic branch of BiSeNet V2 for 8- and 16-fold downsampling, which helped reassign deep feature-map weights and enhanced the ability to extract local key features. Second, a ghost feature-mapping unit was used to optimize the traditional convolution layers and further reduce the computational cost. Third, a new loss function based on the dynamic threshold loss function was applied to supervise the training by adjusting the training weights of the high-loss difficult samples, which enhanced the model's attention to small goals. In experiments on the IDRiD dataset, we conducted an ablation study to verify the effectiveness of each component and compared the proposed model, BiSeNet V2-Pro, with several state-of-the-art models. In comparison with the baseline BiSeNet V2, the segmentation performance of BiSeNet V2-Pro improved by 12.17 %, 11.44 %, and 8.49 % in terms of Sensitivity (SEN), Intersection over Union (IoU), and Dice coefficient (DICE), respectively. Specifically, IoU of MA reaches 0.5716. Compared with other methods, the segmentation speed was significantly improved while ensuring segmentation accuracy, and the number of model parameters was lower. These results demonstrate the superiority of BiSeNet V2-Pro in the multi-lesion segmentation of diabetic retinopathy.

Micro-segmentation of retinal image lesions in diabetic retinopathy using energy-based fuzzy C-Means clustering (EFM-FCM)

Diabetic retinopathy (DR) is a prevalent cause of global visual impairment, contributing to approximately 4.8% of blindness cases worldwide as reported by the World Health Organization (WHO). The condition is characterized by pathological abnormalities in the retinal layer, including microaneurysms, vitreous hemorrhages, and exudates. Microscopic analysis of retinal images is crucial in diagnosing and treating DR. This article proposes a novel method for early DR screening using segmentation and unsupervised learning techniques. The approach integrates a neural network energy-based model into the Fuzzy C-Means (FCM) algorithm to enhance convergence criteria, aiming to improve the accuracy and efficiency of automated DR screening tools. The evaluation of results includes the primary dataset from the Shiva Netralaya Centre, IDRiD, and DIARETDB1. The performance of the proposed method is compared against FCM, EFCM, FLICM, and M-FLICM techniques, utilizing metrics such as accuracy in noiseless and noisy conditions and average execution time. The results showcase auspicious performance on both primary and secondary datasets, achieving accuracy rates of 99.03% in noiseless conditions and 93.13% in noisy images, with an average execution time of 16.1 s. The proposed method holds significant potential in medical image analysis and could pave the way for future advancements in automated DR diagnosis and management. RESEARCH HIGHLIGHTS: A novel approach is proposed in the article, integrating a neural network energy-based model into the FCM algorithm to enhance the convergence criteria and the accuracy of automated DR screening tools. By leveraging the microscopic characteristics of retinal images, the proposed method significantly improves the accuracy of lesion segmentation, facilitating early detection and monitoring of DR. The evaluation of the method's performance includes primary datasets from reputable sources such as the Shiva Netralaya Centre, IDRiD, and DIARETDB1, demonstrating its effectiveness in comparison to other techniques (FCM, EFCM, FLICM, and M-FLICM) in terms of accuracy in both noiseless and noisy conditions. It achieves impressive accuracy rates of 99.03% in noiseless conditions and 93.13% in noisy images, with an average execution time of 16.1 s.

An Improved Microaneurysm Detection Model Based on SwinIR and YOLOv8

Diabetic retinopathy (DR) is a microvascular complication of diabetes. Microaneurysms (MAs) are often observed in the retinal vessels of diabetic patients and represent one of the earliest signs of DR. Accurate and efficient detection of MAs is crucial for the diagnosis of DR. In this study, an automatic model (MA-YOLO) is proposed for MA detection in fluorescein angiography (FFA) images. To obtain detailed features and improve the discriminability of MAs in FFA images, SwinIR was utilized to reconstruct super-resolution images. To solve the problems of missed detection of small features and feature information loss, an MA detection layer was added between the neck and the head sections of YOLOv8. To enhance the generalization ability of the MA-YOLO model, transfer learning was conducted between high-resolution images and low-resolution images. To avoid excessive penalization due to geometric factors and address sample distribution imbalance, the loss function was optimized by taking the Wise-IoU loss as a bounding box regression loss. The performance of the MA-YOLO model in MA detection was compared with that of other state-of-the-art models, including SSD, RetinaNet, YOLOv5, YOLOX, and YOLOv7. The results showed that the MA-YOLO model had the best performance in MA detection, as shown by its optimal metrics, including recall, precision, F1 score, and AP, which were 88.23%, 97.98%, 92.85%, and 94.62%, respectively. Collectively, the proposed MA-YOLO model is suitable for the automatic detection of MAs in FFA images, which can assist ophthalmologists in the diagnosis of the progression of DR.

An effective and comprehensible method to detect and evaluate retinal damage due to diabetes complications

The leading cause of vision loss globally is diabetic retinopathy. Researchers are making great efforts to automatically detect and diagnose correctly diabetic retinopathy. Diabetic retinopathy includes five stages: no diabetic retinopathy, mild diabetic retinopathy, moderate diabetic retinopathy, severe diabetic retinopathy and proliferative diabetic retinopathy. Recent studies have offered several multi-tasking deep learning models to detect and assess the level of diabetic retinopathy. However, the explanation for the assessment of disease severity of these models is limited, and only stops at showing lesions through images. These studies have not explained on what basis the appraisal of disease severity is based. In this article, we present a system for assessing and interpreting the five stages of diabetic retinopathy. The proposed system is built from internal models including a deep learning model that detects lesions and an explanatory model that assesses disease stage. The deep learning model that detects lesions uses the Mask R-CNN deep learning network to specify the location and shape of the lesion and classify the lesion types. This model is a combination of two networks: one used to detect hemorrhagic and exudative lesions, and one used to detect vascular lesions like aneurysm and proliferation. The explanatory model appraises disease severity based on the severity of each type of lesion and the association between types. The severity of the disease will be decided by the model based on the number of lesions, the density and the area of the lesions. The experimental results on real-world datasets show that our proposed method achieves high accuracy of assessing five stages of diabetic retinopathy comparable to existing state-of-the-art methods and is capable of explaining the causes of disease severity.

SSMD-UNet: semi-supervised multi-task decoders network for diabetic retinopathy segmentation

Diabetic retinopathy (DR) is a diabetes complication that can cause vision loss among patients due to damage to blood vessels in the retina. Early retinal screening can avoid the severe consequences of DR and enable timely treatment. Nowadays, researchers are trying to develop automated deep learning-based DR segmentation tools using retinal fundus images to help Ophthalmologists with DR screening and early diagnosis. However, recent studies are unable to design accurate models due to the unavailability of larger training data with consistent and fine-grained annotations. To address this problem, we propose a semi-supervised multitask learning approach that exploits widely available unlabelled data (i.e., Kaggle-EyePACS) to improve DR segmentation performance. The proposed model consists of novel multi-decoder architecture and involves both unsupervised and supervised learning phases. The model is trained for the unsupervised auxiliary task to effectively learn from additional unlabelled data and improve the performance of the primary task of DR segmentation. The proposed technique is rigorously evaluated on two publicly available datasets (i.e., FGADR and IDRiD) and results show that the proposed technique not only outperforms existing state-of-the-art techniques but also exhibits improved generalisation and robustness for cross-data evaluation.

MSLF-Net: A Multi-Scale and Multi-Level Feature Fusion Net for Diabetic Retinopathy Segmentation

Diabetic Retinopathy (DR) is a diabetic complication that predisposes patients to visual impairments that could lead to blindness. Lesion segmentation using deep learning algorithms is an effective measure to screen and prevent early DR. However, there are several types of DR with varying sizes and high inter-class similarity, making segmentation difficult. In this paper, we propose a supervised segmentation method (MSLF-Net) based on multi-scale-multi-level feature fusion to achieve accurate end-to-end DR lesion segmentation. MSLF-Net builds a Multi-Scale Feature Extraction (MSFE) module to extract multi-scale information and provide more comprehensive features for segmentation. This paper further introduces the Multi-Level Feature Fusion (MLFF) module to improve feature fusion using a cross-layer structure. This structure only fuses low- and high-level features of the same class based on category supervision, avoiding feature contamination. Moreover, this paper produces additional masked images for the dataset and performs image enhancement operations to ensure that the proposed method is trainable and functional on small datasets. The extensive experiments are conducted on public datasets IDRID and e_ophtha. The results showed that our proposed feature enhancement method can perform feature fusion more effectively. Therefore, In the end-to-end DR segmentation neural network model, MSLF Net is superior to other similar models in segmentation, and can effectively improve the DR lesion segmentation performance.

ADAM Challenge: Detecting Age-Related Macular Degeneration From Fundus Images

Age-related macular degeneration (AMD) is the leading cause of visual impairment among elderly in the world. Early detection of AMD is of great importance, as the vision loss caused by this disease is irreversible and permanent. Color fundus photography is the most cost-effective imaging modality to screen for retinal disorders. Cutting edge deep learning based algorithms have been recently developed for automatically detecting AMD from fundus images. However, there are still lack of a comprehensive annotated dataset and standard evaluation benchmarks. To deal with this issue, we set up the Automatic Detection challenge on Age-related Macular degeneration (ADAM), which was held as a satellite event of the ISBI 2020 conference. The ADAM challenge consisted of four tasks which cover the main aspects of detecting and characterizing AMD from fundus images, including detection of AMD, detection and segmentation of optic disc, localization of fovea, and detection and segmentation of lesions. As part of the ADAM challenge, we have released a comprehensive dataset of 1200 fundus images with AMD diagnostic labels, pixel-wise segmentation masks for both optic disc and AMD-related lesions (drusen, exudates, hemorrhages and scars, among others), as well as the coordinates corresponding to the location of the macular fovea. A uniform evaluation framework has been built to make a fair comparison of different models using this dataset. During the ADAM challenge, 610 results were submitted for online evaluation, with 11 teams finally participating in the onsite challenge. This paper introduces the challenge, the dataset and the evaluation methods, as well as summarizes the participating methods and analyzes their results for each task. In particular, we observed that the ensembling strategy and the incorporation of clinical domain knowledge were the key to improve the performance of the deep learning models.

Anomaly segmentation in retinal images with poisson-blending data augmentation

Diabetic retinopathy (DR) is one of the most important complications of diabetes. Accurate segmentation of DR lesions is of great importance for the early diagnosis of DR. However, simultaneous segmentation of multi-type DR lesions is technically challenging because of 1) the lack of pixel-level annotations and 2) the large diversity between different types of DR lesions. In this study, first, we propose a novel Poisson-blending data augmentation (PBDA) algorithm to generate synthetic images, which can be easily utilized to expand the existing training data for lesion segmentation. We perform extensive experiments to recognize the important attributes in the PBDA algorithm. We show that position constraints are of great importance and that the synthesis density of one type of lesion has a joint influence on the segmentation of other types of lesions. Second, we propose a convolutional neural network architecture, named DSR-U-Net++ (i.e., DC-SC residual U-Net++), for the simultaneous segmentation of multi-type DR lesions. Ablation studies showed that the mean area under precision recall curve (AUPR) for all four types of lesions increased by >5% with PBDA. The proposed DSR-U-Net++ with PBDA outperformed the state-of-the-art methods by 1.7%-9.9% on the Indian Diabetic Retinopathy Image Dataset (IDRiD) and 67.3% on the e-ophtha dataset with respect to mean AUPR. The developed method would be an efficient tool to generate large-scale task-specific training data for other medical anomaly segmentation tasks.

RTNet: Relation Transformer Network for Diabetic Retinopathy Multi-Lesion Segmentation

Automatic diabetic retinopathy (DR) lesions segmentation makes great sense of assisting ophthalmologists in diagnosis. Although many researches have been conducted on this task, most prior works paid too much attention to the designs of networks instead of considering the pathological association for lesions. Through investigating the pathogenic causes of DR lesions in advance, we found that certain lesions are closed to specific vessels and present relative patterns to each other. Motivated by the observation, we propose a relation transformer block (RTB) to incorporate attention mechanisms at two main levels: a self-attention transformer exploits global dependencies among lesion features, while a cross-attention transformer allows interactions between lesion and vessel features by integrating valuable vascular information to alleviate ambiguity in lesion detection caused by complex fundus structures. In addition, to capture the small lesion patterns first, we propose a global transformer block (GTB) which preserves detailed information in deep network. By integrating the above blocks of dual-branches, our network segments the four kinds of lesions simultaneously. Comprehensive experiments on IDRiD and DDR datasets well demonstrate the superiority of our approach, which achieves competitive performance compared to state-of-the-arts.

LightEyes: A Lightweight Fundus Segmentation Network for Mobile Edge Computing

Fundus is the only structure that can be observed without trauma to the human body. By analyzing color fundus images, the diagnosis basis for various diseases can be obtained. Recently, fundus image segmentation has witnessed vast progress with the development of deep learning. However, the improvement of segmentation accuracy comes with the complexity of deep models. As a result, these models show low inference speeds and high memory usages when deploying to mobile edges. To promote the deployment of deep fundus segmentation models to mobile devices, we aim to design a lightweight fundus segmentation network. Our observation comes from the fact that high-resolution representations could boost the segmentation of tiny fundus structures, and the classification of small fundus structures depends more on local features. To this end, we propose a lightweight segmentation model called LightEyes. We first design a high-resolution backbone network to learn high-resolution representations, so that the spatial relationship between feature maps can be always retained. Meanwhile, considering high-resolution features means high memory usage; for each layer, we use at most 16 convolutional filters to reduce memory usage and decrease training difficulty. LightEyes has been verified on three kinds of fundus segmentation tasks, including the hard exudate, the microaneurysm, and the vessel, on five publicly available datasets. Experimental results show that LightEyes achieves highly competitive segmentation accuracy and segmentation speed compared with state-of-the-art fundus segmentation models, while running at 1.6 images/s Cambricon-1A speed and 51.3 images/s GPU speed with only 36k parameters.

Weakly supervised training for eye fundus lesion segmentation in patients with diabetic retinopathy

OBJECTIVE

Diabetic retinopathy is the leading cause of vision loss in working-age adults. Early screening and diagnosis can help to facilitate subsequent treatment and prevent vision loss. Deep learning has been applied in various fields of medical identification. However, current deep learning-based lesion segmentation techniques rely on a large amount of pixel-level labeled ground truth data, which limits their performance and application. In this work, we present a weakly supervised deep learning framework for eye fundus lesion segmentation in patients with diabetic retinopathy.

METHODS

First, an efficient segmentation algorithm based on grayscale and morphological features is proposed for rapid coarse segmentation of lesions. Then, a deep learning model named Residual-Attention Unet (RAUNet) is proposed for eye fundus lesion segmentation. Finally, a data sample of fundus images with labeled lesions and unlabeled images with coarse segmentation results is jointly used to train RAUNet to broaden the diversity of lesion samples and increase the robustness of the segmentation model.

RESULTS

A dataset containing 582 fundus images with labels verified by doctors, including hemorrhage (HE), microaneurysm (MA), hard exudate (EX) and soft exudate (SE), and 903 images without labels was used to evaluate the model. In ablation test, the proposed RAUNet achieved the highest intersection over union (IOU) on the labeled dataset, and the proposed attention and residual modules both improved the IOU of the UNet benchmark. Using both the images labeled by doctors and the proposed coarse segmentation method, the weakly supervised framework based on RAUNet architecture significantly improved the mean segmentation accuracy by over 7% on the lesions.

SIGNIFICANCE

This study demonstrates that combining unlabeled medical images with coarse segmentation results can effectively improve the robustness of the lesion segmentation model and proposes a practical framework for improving the performance of medical image segmentation given limited labeled data samples.

CARNet: Cascade attentive RefineNet for multi-lesion segmentation of diabetic retinopathy images

Diabetic retinopathy is the leading cause of blindness in working population. Lesion segmentation from fundus images helps ophthalmologists accurately diagnose and grade of diabetic retinopathy. However, the task of lesion segmentation is full of challenges due to the complex structure, the various sizes and the interclass similarity with other fundus tissues. To address the issue, this paper proposes a cascade attentive RefineNet (CARNet) for automatic and accurate multi-lesion segmentation of diabetic retinopathy. It can make full use of the fine local details and coarse global information from the fundus image. CARNet is composed of global image encoder, local image encoder and attention refinement decoder. We take the whole image and the patch image as the dual input, and feed them to ResNet50 and ResNet101, respectively, for downsampling to extract lesion features. The high-level refinement decoder uses dual attention mechanism to integrate the same-level features in the two encoders with the output of the low-level attention refinement module for multiscale information fusion, which focus the model on the lesion area to generate accurate predictions. We evaluated the segmentation performance of the proposed CARNet on the IDRiD, E-ophtha and DDR data sets. Extensive comparison experiments and ablation studies on various data sets demonstrate the proposed framework outperforms the state-of-the-art approaches and has better accuracy and robustness. It not only overcomes the interference of similar tissues and noises to achieve accurate multi-lesion segmentation, but also preserves the contour details and shape features of small lesions without overloading GPU memory usage.

Pyramid-Net: Intra-layer Pyramid-Scale Feature Aggregation Network for Retinal Vessel Segmentation

Retinal vessel segmentation plays an important role in the diagnosis of eye-related diseases and biomarkers discovery. Existing works perform multi-scale feature aggregation in an inter-layer manner, namely inter-layer feature aggregation. However, such an approach only fuses features at either a lower scale or a higher scale, which may result in a limited segmentation performance, especially on thin vessels. This discovery motivates us to fuse multi-scale features in each layer, intra-layer feature aggregation, to mitigate the problem. Therefore, in this paper, we propose Pyramid-Net for accurate retinal vessel segmentation, which features intra-layer pyramid-scale aggregation blocks (IPABs). At each layer, IPABs generate two associated branches at a higher scale and a lower scale, respectively, and the two with the main branch at the current scale operate in a pyramid-scale manner. Three further enhancements including pyramid inputs enhancement, deep pyramid supervision, and pyramid skip connections are proposed to boost the performance. We have evaluated Pyramid-Net on three public retinal fundus photography datasets (DRIVE, STARE, and CHASE-DB1). The experimental results show that Pyramid-Net can effectively improve the segmentation performance especially on thin vessels, and outperforms the current state-of-the-art methods on all the adopted three datasets. In addition, our method is more efficient than existing methods with a large reduction in computational cost. We have released the source code at https://github.com/JerRuy/Pyramid-Net.

A Magnified Adaptive Feature Pyramid Network for automatic microaneurysms detection

Diabetic retinopathy (DR), as an important complication of diabetes, is the primary cause of blindness in adults. Automatic DR detection poses a challenge which is crucial for early DR screening. Currently, the vast majority of DR is diagnosed through fundus images, where the microaneurysm (MA) has been widely used as the most distinguishable marker. Research works on automatic DR detection have traditionally utilized manually designed operators, while a few recent researchers have explored deep learning techniques for this topic. But due to issues such as the extremely small size of microaneurysms, low resolution of fundus pictures, and insufficient imaging depth, the DR detection problem is quite challenging and remains unsolved. To address these issues, this research proposes a new deep learning model (Magnified Adaptive Feature Pyramid Network, MAFP-Net) for DR detection, which conducts super-resolution on low quality fundus images and integrates an improved feature pyramid structure while utilizing a standard two-stage detection network as the backbone. Our proposed detection model needs no pre-segmented patches to train the CNN network. When tested on the E-ophtha-MA dataset, the sensitivity value of our method reached as high as 83.5% at false positives per image (FPI) of 8 and the F1 value achieved 0.676, exceeding all those of the state-of-the-art algorithms as well as the human performance of experienced physicians. Similar results were achieved on another public dataset of IDRiD.

Fundus image segmentation via hierarchical feature learning

Fundus Image Segmentation (FIS) is an essential procedure for the automated diagnosis of ophthalmic diseases. Recently, deep fully convolutional networks have been widely used for FIS with state-of-the-art performance. The representative deep model is the U-Net, which follows an encoder-decoder architecture. I believe it is suboptimal for FIS because consecutive pooling operations in the encoder lead to low-resolution representation and loss of detailed spatial information, which is particularly important for the segmentation of tiny vessels and lesions. Motivated by this, a high-resolution hierarchical network (HHNet) is proposed to learn semantic-rich high-resolution representations and preserve spatial details simultaneously. Specifically, a High-resolution Feature Learning (HFL) module with increasing dilation rates was first designed to learn the high-level high-resolution representations. Then, the HHNet was constructed by incorporating three HFL modules and two feature aggregation modules. The HHNet runs in a coarse-to-fine manner, and fine segmentation maps are output at the last level. Extensive experiments were conducted on fundus lesion segmentation, vessel segmentation, and optic cup segmentation. The experimental results reveal that the proposed method shows highly competitive or even superior performance in terms of segmentation performance and computation cost, indicating its potential advantages in clinical application.

EAD-Net: A Novel Lesion Segmentation Method in Diabetic Retinopathy Using Neural Networks

Diabetic retinopathy (DR) is a common chronic fundus disease, which has four different kinds of microvessel structure and microvascular lesions: microaneurysms (MAs), hemorrhages (HEs), hard exudates, and soft exudates. Accurate detection and counting of them are a basic but important work. The manual annotation of these lesions is a labor-intensive task in clinical analysis. To solve the problem, we proposed a novel segmentation method for different lesions in DR. Our method is based on a convolutional neural network and can be divided into encoder module, attention module, and decoder module, so we refer it as EAD-Net. After normalization and augmentation, the fundus images were sent to the EAD-Net for automated feature extraction and pixel-wise label prediction. Given the evaluation metrics based on the matching degree between detected candidates and ground truth lesions, our method achieved sensitivity of 92.77%, specificity of 99.98%, and accuracy of 99.97% on the e_ophtha_EX dataset and comparable AUPR (Area under Precision-Recall curve) scores on IDRiD dataset. Moreover, the results on the local dataset also show that our EAD-Net has better performance than original U-net in most metrics, especially in the sensitivity and F1-score, with nearly ten percent improvement. The proposed EAD-Net is a novel method based on clinical DR diagnosis. It has satisfactory results on the segmentation of four different kinds of lesions. These effective segmentations have important clinical significance in the monitoring and diagnosis of DR.

Dual-Branch Network with Dual-Sampling Modulated Dice Loss for Hard Exudate Segmentation in Colour Fundus Images

Automated segmentation of hard exudates in colour fundus images is a challenge task due to issues of extreme class imbalance and enormous size variation. This paper aims to tackle these issues and proposes a dual-branch network with dual-sampling modulated Dice loss. It consists of two branches: large hard exudate biased segmentation branch and small hard exudate biased segmentation branch. Both of them are responsible for their own duties separately. Furthermore, we propose a dual-sampling modulated Dice loss for the training such that our proposed dual-branch network is able to segment hard exudates in different sizes. In detail, for the first branch, we use a uniform sampler to sample pixels from predicted segmentation mask for Dice loss calculation, which leads to this branch naturally be biased in favour of large hard exudates as Dice loss generates larger cost on misidentification of large hard exudates than small hard exudates. For the second branch, we use a re-balanced sampler to oversample hard exudate pixels and undersample background pixels for loss calculation. In this way, cost on misidentification of small hard exudates is enlarged, which enforces the parameters in the second branch fit small hard exudates well. Considering that large hard exudates are much easier to be correctly identified than small hard exudates, we propose an easy-to-difficult learning strategy by adaptively modulating the losses of two branches. We evaluate our proposed method on two public datasets and results demonstrate that ours achieves state-of-the-art performance.

Deep Bayesian baseline for segmenting diabetic retinopathy lesions: Advances and challenges

Early diagnosis of retinopathy is essential for preventing retinal complications and visual impairment due to diabetes. For the detection of retinopathy lesions from retinal images, several automatic approaches based on deep neural networks have been developed in the recent years. Most of the proposed methods produce point estimates of pixels belonging to the lesion areas and give no or little information on the uncertainty of method predictions. However, the latter can be essential in the examination of the medical condition of the patient when the goal is early detection of abnormalities. This work extends the recent research with a Bayesian framework by considering the parameters of a convolutional neural network as random variables and utilizing stochastic variational dropout based approximation for uncertainty quantification. The framework includes an extended validation procedure and it allows analyzing lesion segmentation distributions, model calibration and prediction uncertainties. Also the challenges related to the deep probabilistic model and uncertainty quantification are presented. The proposed method achieves area under precision-recall curve of 0.84 for hard exudates, 0.641 for soft exudates, 0.593 for haemorrhages, and 0.484 for microaneurysms on IDRiD dataset.

Deep learning for diabetic retinopathy detection and classification based on fundus images: A review

Diabetic Retinopathy is a retina disease caused by diabetes mellitus and it is the leading cause of blindness globally. Early detection and treatment are necessary in order to delay or avoid vision deterioration and vision loss. To that end, many artificial-intelligence-powered methods have been proposed by the research community for the detection and classification of diabetic retinopathy on fundus retina images. This review article provides a thorough analysis of the use of deep learning methods at the various steps of the diabetic retinopathy detection pipeline based on fundus images. We discuss several aspects of that pipeline, ranging from the datasets that are widely used by the research community, the preprocessing techniques employed and how these accelerate and improve the models' performance, to the development of such deep learning models for the diagnosis and grading of the disease as well as the localization of the disease's lesions. We also discuss certain models that have been applied in real clinical settings. Finally, we conclude with some important insights and provide future research directions.

Applications of deep learning in fundus images: A review

The use of fundus images for the early screening of eye diseases is of great clinical importance. Due to its powerful performance, deep learning is becoming more and more popular in related applications, such as lesion segmentation, biomarkers segmentation, disease diagnosis and image synthesis. Therefore, it is very necessary to summarize the recent developments in deep learning for fundus images with a review paper. In this review, we introduce 143 application papers with a carefully designed hierarchy. Moreover, 33 publicly available datasets are presented. Summaries and analyses are provided for each task. Finally, limitations common to all tasks are revealed and possible solutions are given. We will also release and regularly update the state-of-the-art results and newly-released datasets at https://github.com/nkicsl/Fundus_Review to adapt to the rapid development of this field.

An efficient framework for automated screening of Clinically Significant Macular Edema

The present study proposes a new approach to automated screening of Clinically Significant Macular Edema (CSME) and addresses two major challenges associated with such screenings, i.e., exudate segmentation and imbalanced datasets. The proposed approach replaces the conventional exudate segmentation based feature extraction by combining a pre-trained deep neural network with meta-heuristic feature selection. A feature space over-sampling technique is being used to overcome the effects of skewed datasets and the screening is accomplished by a k-NN based classifier. The role of each data-processing step (e.g., class balancing, feature selection) and the effects of limiting the region of interest to fovea on the classification performance are critically analyzed. Finally, the selection and implication of operating points on Receiver Operating Characteristic curve are discussed. The results of this study convincingly demonstrate that by following these fundamental practices of machine learning, a basic k-NN based classifier could effectively accomplish the CSME screening.

Automatic Drusen Segmentation for Age-Related Macular Degeneration in Fundus Images Using Deep Learning

Drusen are the main aspect of detecting age-related macular degeneration (AMD). Ophthalmologists can evaluate the condition of AMD based on drusen in fundus images. However, in the early stage of AMD, the drusen areas are usually small and vague. This leads to challenges in the drusen segmentation task. Moreover, due to the high-resolution fundus images, it is hard to accurately predict the drusen areas with deep learning models. In this paper, we propose a multi-scale deep learning model for drusen segmentation. By exploiting both local and global information, we can improve the performance, especially in the early stages of AMD cases.

AVNet: A retinal artery/vein classification network with category-attention weighted fusion

BACKGROUND AND OBJECTIVE

Automatic artery/vein (A/V) classification in retinal images is of great importance in detecting vascular abnormalities, which may provide biomarkers for early diagnosis of many systemic diseases. It is intuitive to apply popular deep semantic segmentation network for A/V classification. However, the model is required to provide powerful representation ability since vessel is much more complex than general objects. Moreover, deep network may lead to inconsistent classification results for the same vessel due to the lack of structured optimization objective.

METHODS

In this paper, we propose a novel segmentation network named AVNet, which effectively enhances the classification ability of the model by integrating category-attention weighted fusion (CWF) module, significantly improving the pixel-level A/V classification results. Then, a graph based vascular structure reconstruction (VSR) algorithm is employed to reduce the segment-wise inconsistency, verifying the effect of the graph model on noisy vessel segmentation results.

RESULTS

The proposed method has been verified on three datasets, i.e. DRIVE, LES-AV and WIDE. AVNet achieves pixel-level accuracies of 90.62%, 90.34%, and 93.16%, respectively, and VSR further improves the performance by 0.19%, 1.85% and 0.64%, achieving the state-of-the-art results on these three datasets.

CONCLUSION

The proposed method achieves competitive performance in A/V classification task.