Retinal Vessels In Fundus Images Research Articles

RLeU-Net: Segmentation of blood vessels in retinal fundus images for Diabetic Retinopathy Screening

RLeU-Net: Segmentation of blood vessels in retinal fundus images for Diabetic Retinopathy Screening

High-precision retinal blood vessel segmentation based on a multi-stage and dual-channel deep learning network.

Objective.The high-precision segmentation of retinal vessels in fundus images is important for the early diagnosis of ophthalmic diseases. However, the extraction for microvessels is challenging due to their characteristics of low contrast and high structural complexity. Although some works have been developed to improve the segmentation ability in thin vessels, they have only been successful in recognizing small vessels with relatively high contrast.Approach.Therefore, we develop a deep learning (DL) framework with a multi-stage and dual-channel network model (MSDC_NET) to further improve the thin-vessel segmentation with low contrast. Specifically, an adaptive image enhancement strategy combining multiple preprocessing and the DL method is firstly proposed to elevate the contrast of thin vessels; then, a two-channel model with multi-scale perception is developed to implement whole- and thin-vessel segmentation; and finally, a series of post-processing operations are designed to extract more small vessels in the predicted maps from thin-vessel channels.Main results.Experiments on DRIVE, STARE and CHASE_DB1 demonstrate the superiorities of the proposed MSDC_NET in extracting more thin vessels in fundus images, and quantitative evaluations on several parameters based on the advanced ground truth further verify the advantages of our proposed DL model. Compared with the previous multi-branch method, the specificity and F1score are improved by about 2.18%, 0.68%, 1.73% and 2.91%, 0.24%, 8.38% on the three datasets, respectively.Significance.This work may provide richer information to ophthalmologists for the diagnosis and treatment of vascular-related ophthalmic diseases.

EAMR-Net: A multiscale effective spatial and cross-channel attention network for retinal vessel segmentation.

Delineation of retinal vessels in fundus images is essential for detecting a range of eye disorders. An automated technique for vessel segmentation can assist clinicians and enhance the efficiency of the diagnostic process. Traditional methods fail to extract multiscale information, discard unnecessary information, and delineate thin vessels. In this paper, a novel residual U-Net architecture that incorporates multi-scale feature learning and effective attention is proposed to delineate the retinal vessels precisely. Since drop block regularization performs better than drop out in preventing overfitting, drop block was used in this study. A multi-scale feature learning module was added instead of a skip connection to learn multi-scale features. A novel effective attention block was proposed and integrated with the decoder block to obtain precise spatial and channel information. Experimental findings indicated that the proposed model exhibited outstanding performance in retinal vessel delineation. The sensitivities achieved for DRIVE, STARE, and CHASE_DB datasets were 0.8293, 0.8151 and 0.8084, respectively.

Segmentation of retinal vessels in fundus images based on U-Net with self-calibrated convolutions and spatial attention modules.

Automated and accurate segmentation of retinal vessels in fundus images is an important step for screening and diagnosing various ophthalmologic diseases. However, many factors, including the variations of vessels in color, shape and size, make this task become an intricate challenge. One kind of the most popular methods for vessel segmentation is U-Net based methods. However, in the U-Net based methods, the size of the convolution kernels is generally fixed. As a result, the receptive field for an individual convolution operation is single, which is not conducive to the segmentation of retinal vessels with various thicknesses. To overcome this problem, in this paper, we employed self-calibrated convolutions to replace the traditional convolutions for the U-Net, which can make the U-Net learn discriminative representations from different receptive fields. Besides, we proposed an improved spatial attention module, instead of using traditional convolutions, to connect the encoding part and decoding part of the U-Net, which can improve the ability of the U-Net to detect thin vessels. The proposed method has been tested on Digital Retinal Images for Vessel Extraction (DRIVE) database and Child Heart and Health Study in England Database (CHASE DB1). The metrics used to evaluate the performance of the proposed method are accuracy (ACC), sensitivity (SE), specificity (SP), F1-score (F1) and the area under the receiver operating characteristic curve (AUC). The ACC, SE, SP, F1 and AUC obtained by the proposed method are 0.9680, 0.8036, 0.9840, 0.8138 and 0.9840 respectively on DRIVE database, and 0.9756, 0.8118, 0.9867, 0.8068 and 0.9888 respectively on CHASE DB1, which are better than those obtained by the traditional U-Net (the ACC, SE, SP, F1 and AUC obtained by U-Net are 0.9646, 0.7895, 0.9814, 0.7963 and 0.9791 respectively on DRIVE database, and 0.9733, 0.7817, 0.9862, 0.7870 and 0.9810 respectively on CHASE DB1). The experimental results indicate that the proposed modifications in the U-Net are effective for vessel segmentation. The structure of the proposed network.

A lightweight dual-path cascaded network for vessel segmentation in fundus image.

Automatic and fast segmentation of retinal vessels in fundus images is a prerequisite in clinical ophthalmic diseases; however, the high model complexity and low segmentation accuracy still limit its application. This paper proposes a lightweight dual-path cascaded network (LDPC-Net) for automatic and fast vessel segmentation. We designed a dual-path cascaded network via two U-shaped structures. Firstly, we employed a structured discarding (SD) convolution module to alleviate the over-fitting problem in both codec parts. Secondly, we introduced the depthwise separable convolution (DSC) technique to reduce the parameter amount of the model. Thirdly, a residual atrous spatial pyramid pooling (ResASPP) model is constructed in the connection layer to aggregate multi-scale information effectively. Finally, we performed comparative experiments on three public datasets. Experimental results show that the proposed method achieved superior performance on the accuracy, connectivity, and parameter quantity, thus proving that it can be a promising lightweight assisted tool for ophthalmic diseases.

Intelligent Machine Learning Enabled Retinal Blood Vessel Segmentation and Classification

Automated segmentation of blood vessels in retinal fundus images is essential for medical image analysis. The segmentation of retinal vessels is assumed to be essential to the progress of the decision support system for initial analysis and treatment of retinal disease. This article develops a new Grasshopper Optimization with Fuzzy Edge Detection based Retinal Blood Vessel Segmentation and Classification (GOFED-RBVSC) model. The proposed GOFED-RBVSC model initially employs contrast enhancement process. Besides, GOAFED approach is employed to detect the edges in the retinal fundus images in which the use of GOA adjusts the membership functions. The ORB (Oriented FAST and Rotated BRIEF) feature extractor is exploited to generate feature vectors. Finally, Improved Conditional Variational Auto Encoder (ICAVE) is utilized for retinal image classification, shows the novelty of the work. The performance validation of the GOFED-RBVSC model is tested using benchmark dataset, and the comparative study highlighted the betterment of the GOFED-RBVSC model over the recent approaches.

Fuzzy based image edge detection algorithm for blood vessel detection in retinal images

We developed a contour detection based image processing algorithm based on Mamdani (Type-2) fuzzy rules for detection of blood vessels in retinal fundus images. The method uses the green channel data from eye fundus images as input, Contrast-Limited Adaptive Histogram Equalization (CLAHE) for contrast enhancement, and median filter for background exclusion. The Mamdani (Type-2) fuzzy rules applied on image gradient value are used for edge detection. The results of experiments on the Digital Retinal Images for Vessel Extraction (DRIVE), STructured Analysis of the Retina (STARE) and CHASEdb datasets show the applicability of the proposed method as a flexible approach which can be adapted to numerous edge detection/contour based applications. We achieved an accuracy of 0.865 for STARE dataset, an accuracy of 0.939 for the DRIVE dataset, and the accuracy of 0.950 for the ChaseDB dataset. In relation to works of other authors, our method offered a similar performance, but it offers an improved dynamics and flexibility in formulation of the linguistic threshold criteria, which can be a leading factor in design of image processing systems with dynamic and flexible rules, such as Type-2 fuzzy rules would allow, offering an interesting alternative to currently widespread deep learning applications.

Multimodal Registration of Fundus Images With Fluorescein Angiography for Fine-Scale Vessel Segmentation

Background and objective: The analysis of retinal vessels in fundus images is vital in the diagnosis of retinal diseases and early diagnosis of chronic vascular diseases and diabetes. Automatic vessel segmentation relies on costly expert annotations that may have limitations in the level of detail. We develop an automatic method to generate highly-accurate vessel segmentation including fine-scale vessels. Methods: We present a new framework for fine-scale vessel segmentation from fundus images through registration and segmentation of corresponding fluorescein angiography (FA) images. We first register and aggregate the extracted vessels highlighted from the fluorescent dye in the FA frames. This FA vessel mask is then registered to the fundus image based on an initial fundus vessel mask. Post-processing is performed to refine the final vessel mask. Registration of the FA frames, and registration of FA vessel mask to the fundus image, are performed by similar coarse-to-fine hierarchical frameworks comprising both projective and deformable registration. Two convolutional neural networks with identical network structures, both trained on public datasets but with different configurations, are used for vessel segmentation of both the FA frames and the fundus images. Results: Qualitative examples support the robustness and accuracy of the proposed method. Quantitative evaluations, including the area-under-curve (AUC) of the receiver operating characteristic (ROC) curve are presented. Although fair comparisons cannot be made due to a lack of similar methods and adequate public datasets, we demonstrate that the proposed method with an AUC ROC of 0.979, outperforms a state-of-the-art automatic vessel segmentation method trained on publicly available datasets at 0.956. Conclusions: The proposed method generates accurate vessel segmentation results containing filamentary vessels that are virtually indiscernible to the naked eye in color retinal fundus images.

Determinación de la parábola de la vasculatura de la retina mediante un algoritmo computacional de segmentación

El análisis cuantitativo de la arquitectura de las venas temporales superior e inferior y su monitoreo sobre el tiempo puede facilitar el diagnóstico y tratamiento oportuno de la retinopatía diabética. En este trabajo se presenta un novedoso método que consiste de dos etapas correspondientes a la segmentación automática y modelado parabólico de las venas temporales superior e inferior en imágenes de fondo de ojo. En la primera etapa, el detector lineal multiescala (DLM) es empleado para detectar estructuras de tipo arterial en imágenes de la retina. Debido a que DLM es un método de realzado arterial, es necesario aplicar una estrategia de umbralización para clasificar pixeles de tipo arterial con respecto al fondo de la imagen, donde un valor de umbral determinado de forma experimental es comparado con cinco métodos de umbralización del estado del arte. En esta etapa, el método de segmentación propuesto es comparado con seis métodos especializados del estado del arte en términos de eficiencia de segmentación. En la segunda etapa, se desempeña un modelado parabólico mediante una estrategia de optimización utilizando un Algoritmo de Distribución Marginal Univariada sobre las arterias previamente segmentadas, y los resultados son comparados con dos métodos paramétricos del estado del arte y con las delineaciones realizadas por especialistas. Los resultados de segmentación arterial utilizando el detector lineal multiescala demostraron una alta eficiencia de segmentación obteniendo un valor de 0.9618 utilizando la base de datos DRIVE de imágenes de fondo de ojo. De igual forma, los resultados de modelado parabólico entregaron una eficiencia promedio de 0.825 con respecto a las delineaciones realizadas por especialistas oftalmólogos de las venas temporales superior e inferior. En base a los resultados de eficiencia y al tiempo computacional (5.62 segundos), el método propuesto puede considerarse como altamente apropiado para desempeñar diagnóstico asistido por computadora en el área de oftalmología.

GGM classifier with multi-scale line detectors for retinal vessel segmentation

Persistent changes in the diameter of retinal blood vessels may indicate some chronic eye diseases. Computer-assisted change observation attempts may become challenging due to the emergence of interfering pathologies around blood vessels in retinal fundus images. The end result is lower sensitivity to thin vessels for certain computerized detection methods. Quite recently, multi-scale line detection method proved to be worthy for improved sensitivity toward lower-caliber vessels detection. This happens largely due to its adaptive property that responds more to the longevity patterns than width of a given vessel. However, the method suffers from the lack of a better aggregation process for individual line detectors. This paper investigates a scenario that introduces a supervised generalized Gaussian mixture classifier as a robust solution for the aggregate process. The classifier is built with class-conditional probability density functions as a logistic function of linear mixtures. To boost the classifier’s performance, the weighted scale images are modeled as Gaussian mixtures. The classifier is trained with weighted images modeled on a Gaussian mixture. The net effect is increased sensitivity for small vessels. The classifier’s performance has been tested with three commonly available data sets: DRIVE, SATRE, and CHASE_DB1. The results of the proposed method (with an accuracy of 96%, 96.1% and 95% on DRIVE, STARE, and CHASE_DB1, respectively) demonstrate its competitiveness against the state-of-the-art methods and its reliability for vessel segmentation.

DUNet: A deformable network for retinal vessel segmentation

Automatic segmentation of retinal vessels in fundus images plays an important role in the diagnosis of some diseases such as diabetes and hypertension. In this paper, we propose Deformable U-Net (DUNet), which exploits the retinal vessels’ local features with a U-shape architecture, in an end to end manner for retinal vessel segmentation. Inspired by the recently introduced deformable convolutional networks, we integrate the deformable convolution into the proposed network. The DUNet, with upsampling operators to increase the output resolution, is designed to extract context information and enable precise localization by combining low-level features with high-level ones. Furthermore, DUNet captures the retinal vessels at various shapes and scales by adaptively adjusting the receptive fields according to vessels’ scales and shapes. Public datasets: DRIVE, STARE, CHASE_DB1 and HRF are used to test our models. Detailed comparisons between the proposed network and the deformable neural network, U-Net are provided in our study. Results show that more detailed vessels can be extracted by DUNet and it exhibits state-of-the-art performance for retinal vessel segmentation with a global accuracy of 0.9566/0.9641/0.9610/0.9651 and AUC of 0.9802/0.9832/0.9804/0.9831 on DRIVE, STARE, CHASE_DB1 and HRF respectively. Moreover, to show the generalization ability of the DUNet, we use another two retinal vessel data sets, i.e., WIDE and SYNTHE, to qualitatively and quantitatively analyze and compare with other methods. Extensive cross-training evaluations are used to further assess the extendibility of DUNet. The proposed method has the potential to be applied to the early diagnosis of diseases.

Construction of Retinal Vessel Segmentation Models Based on Convolutional Neural Network

Segmentation of retinal vessels in fundus images plays a very important role in diagnosing relevant diseases. In this paper, we have constructed automated segmentation models for the retinal vessel segmentation task based on convolutional neural networks. Since some typical deep convolutional neural networks need to be fed by high-resolution patches, small retinal patches should be interpolated to the specific resolution. The interpolated patches sometimes would introduce additional noises. Thus, we modify some typical deep architectures by inserting a set of convolutional layers. In this way, our models have the ability to adapt to different resolutions. Overall, five models are analyzed and compared in our studies including LeNet, M-AlexNet (modified AlexNet), M-ZF-Net (Modified ZF-Net), M-VGG (Modified VGG) and Deformable-ConvNet. Deformable-ConvNet captures the vascular structure and is used to do the retinal vessel segmentation task for the first time. We train the models from scratch and compare their ability to discriminate vessels/non-vessel pixels on two retinal fundus image datasets, DRIVE and STARE. Results are analyzed and compared in our studies. We obtain the highest accuracy of 0.9628/0.9690, lowest loss of 0.1045/0.0968, and highest AUC of 0.9764/0.9844 on DRIVE/STARE respectively. We also compare the CNN models with other segmentation methods. The results demonstrate the high effectiveness of the CNN-based approaches.

Morphological and Otsu’s Thresholding-Based Retinal Blood Vessel Segmentation for Detection of Retinopathy

Retinal blood vessel segmentation is crucial as it is the earliest process in measuring various indicators of retinopathy sign such as arterial-venous nicking, and focal arteriolar and generalized arteriolar narrowing. The segmentation can be clinically used if its accuracy is close to 100%. In this study, a new method of segmentation is developed for extraction of retinal blood vessel. In this paper, we present a new automated method to extract blood vessels in retinal fundus images. The proposed method comprises of two main parts and a few subcomponents which include pre-processing and segmentation. The main focus for the segmentation part is two morphological reconstructions which are the morphological reconstructions followed by the morphological top-hat transform. Then the technique to classify the vessel pixels and background pixels is Otsu’s Thresholding. The image database used in this study is the High Resolution Fundus Image Database (HRFID). The developed segmentation method accuracies are 95.17%, 92.06% and 94.71% when tested on dataset of healthy, diabetic retinopathy (DR) and glaucoma patients respectively. Overall, the performance of the proposed method is comparable with existing methods with overall accuracies were more than 90 % for all three different categories: healthy, DR and glaucoma.

Artery/vein classification of retinal blood vessels using feature selection.

Automated classification of retinal vessels in fundus images is the first step towards measurement of retinal characteristics that can be used to screen and diagnose vessel abnormalities for cardiovascular and retinal disorders. This paper presents a novel approach to vessel classification to compute the artery/vein ratio (AVR) for all blood vessel segments in the fundus image. The features extracted are then subjected to a selection procedure using Random Forests (RF) where the features that contribute most to classification accuracy are chosen as input to a polynomial kernel Support Vector Machine (SVM) classifier. The most dominant feature was found to be the vessel information obtained from the Light plane of the LAB color space. The SVM is then subjected to one time training using 10-fold cross validation on images randomly selected from the VICAVR dataset before testing on an independent test dataset, derived from the same database. An Area Under the ROC Curve (AUC) of 97.2% was obtained on an average of 100 runs of the algorithm. The proposed algorithm is robust due to the feature selection procedure, and it is possible to get similar accuracies across many datasets.

A simple hybrid method for segmenting vessel structures in retinal\\ fundus images

In this paper, a simple, fast, and efficient hybrid segmentation method is presented for extracting vessel structures in retinal fundus images. Basically, this hybrid approach combines circular and naive Bayes classifiers to extract blood vessels in retinal fundus images. The circular method samples pixels along the enlarging circles centered at the current pixel and classifies the current pixel as vessel or nonvessel. An elimination technique is then employed to eliminate the nonvessel fragments from the processed image. The naive Bayes method as a supervised technique uses a very small set of features to segment retinal vessels in retinal images. The designed hybrid method exploits the circular and Bayesian segmentation results together to achieve the best performance. The achieved performance of the segmentation methods are tested on DRIVE and STARE databases for evaluation. The proposed methods segment a retinal image within 1 s and achieve about 95% accuracy. The results also indicate that the proposed hybrid method is one of the simplest and efficient segmentation methods among the unsupervised and supervised methods in the literature.

Deep neural network and random forest hybrid architecture for learning to detect retinal vessels in fundus images.

Vision impairment due to pathological damage of the retina can largely be prevented through periodic screening using fundus color imaging. However the challenge with large-scale screening is the inability to exhaustively detect fine blood vessels crucial to disease diagnosis. In this work we present a computational imaging framework using deep and ensemble learning based hybrid architecture for reliable detection of blood vessels in fundus color images. A deep neural network (DNN) is used for unsupervised learning of vesselness dictionaries using sparse trained denoising auto-encoders (DAE), followed by supervised learning of the DNN response using a random forest for detecting vessels in color fundus images. In experimental evaluation with the DRIVE database, we achieve the objective of vessel detection with max. avg. accuracy of 0.9327 and area under ROC curve of 0.9195.

An automated computational framework for retinal vascular network labeling and branching order analysis

Changes in retinal vascular morphology are well known as predictive clinical signs of many diseases such as hypertension, diabetes and so on. Computer-aid image processing and analysis for retinal vessels in fundus images are effective and efficient in clinical diagnosis instead of tedious manual labeling and measurement. An automated computational framework for retinal vascular network labeling and analysis is presented in this work. The framework includes 1) detecting and locating the optic disc; 2) tracking the vessel centerline from detected seed points and linking the breaks after tracing; 3) extracting all the retinal vascular trees and identifying all the significant points; and 4) classifying terminal points into starting points and ending points based on the information of optic disc location, and finally assigning branch order for each extracted vascular tree in the image. All the modules in the framework are fully automated. Based on the results, morphological analysis is then applied to achieve geometrical and topological features based on branching order for one individual vascular tree or for the vascular network through the retinal vascular network in the images. Validation and experiments on the public DRIVE database have demonstrated that the proposed framework is a novel approach to analyze and study the vascular network pattern, and may offer new insights to the diagnosis of retinopathy.

Detection and measurement of retinal vessels in fundus images using amplitude modified second-order Gaussian filter

In this paper, the fitness of estimating vessel profiles with Gaussian function is evaluated and an amplitude-modified second-order Gaussian filter is proposed for the detection and measurement of vessels. Mathematical analysis is given and supported by a simulation and experiments to demonstrate that the vessel width can be measured in linear relationship with the "spreading factor" of the matched filter when the magnitude coefficient of the filter is suitably assigned. The absolute value of vessel diameter can be determined simply by using a precalibrated line, which is typically required since images are always system dependent. The experiment shows that the inclusion of the width measurement in the detection process can improve the performance of matched filter and result in a significant increase in success rate of detection.