The accurate segmentation of retinal images is a basic step in screening for retinopathy and glaucoma. Most existing retinal image segmentation methods have insufficient feature information extraction. They are susceptible to the impact of the lesion area and poor image quality, resulting in the poor recovery of contextual information. This also causes the segmentation results of the model to be noisy and low in accuracy. Therefore, this paper proposes a multi-scale and multi-branch convolutional neural network model (multi-scale and multi-branch network (MSMB-Net)) for retinal image segmentation. The model uses atrous convolution with different expansion rates and skip connections to reduce the loss of feature information. Receiving domains of different sizes captures global context information. The model fully integrates shallow and deep semantic information and retains rich spatial information. The network embeds an improved attention mechanism to obtain more detailed information, which can improve the accuracy of segmentation. Finally, the method of this paper was validated on the fundus vascular datasets, DRIVE, STARE and CHASE datasets, with accuracies/F1 of 0.9708/0.8320, 0.9753/0.8469 and 0.9767/0.8190, respectively. The effectiveness of the method in this paper was further validated on the optic disc visual cup DRISHTI-GS1 dataset with an accuracy/F1 of 0.9985/0.9770. Experimental results show that, compared with existing retinal image segmentation methods, our proposed method has good segmentation performance in all four benchmark tests.

Accurate segmentation of retinal blood vessels is a key step in the diagnosis of fundus diseases, among which cataracts, glaucoma, and diabetic retinopathy (DR) are the main diseases that cause blindness. Most segmentation methods based on deep convolutional neural networks can effectively extract features. However, convolution and pooling operations also filter out some useful information, and the final segmented retinal vessels have problems such as low classification accuracy. In this paper, we propose a multi-scale residual attention network called MRA-UNet. Multi-scale inputs enable the network to learn features at different scales, which increases the robustness of the network. In the encoding phase, we reduce the negative influence of the background and eliminate noise by using the residual attention module. We use the bottom reconstruction module to aggregate the feature information under different receptive fields, so that the model can extract the information of different thicknesses of blood vessels. Finally, the spatial activation module is used to process the up-sampled image to further increase the difference between blood vessels and background, which promotes the recovery of small blood vessels at the edges. Our method was verified on the DRIVE, CHASE, and STARE datasets. Respectively, the segmentation accuracy rates reached 96.98%, 97.58%, and 97.63%; the specificity reached 98.28%, 98.54%, and 98.73%; and the F-measure scores reached 82.93%, 81.27%, and 84.22%. We compared the experimental results with some state-of-art methods, such as U-Net, R2U-Net, and AG-UNet in terms of accuracy, sensitivity, specificity, F-measure, and AUC${}_{ROC}$. Particularly, MRA-UNet outperformed U-Net by 1.51%, 3.44%, and 0.49% on DRIVE, CHASE, and STARE datasets, respectively.

Accurate segmentation of retinal vessels is critical to the mechanism, diagnosis, and treatment of many ocular pathologies. Due to the poor contrast and inhomogeneous background of fundus imaging and the complex structure of retinal fundus images, this makes accurate segmentation of blood vessels from retinal images still challenging. In this paper, we propose an effective framework for retinal vascular segmentation, which is innovative mainly in the retinal image pre-processing stage and segmentation stage. First, we perform image enhancement on three publicly available fundus datasets based on the multiscale retinex with color restoration (MSRCR) method, which effectively suppresses noise and highlights the vessel structure creating a good basis for the segmentation phase. The processed fundus images are then fed into an effective Reverse Fusion Attention Residual Network (RFARN) for training to achieve more accurate retinal vessel segmentation. In the RFARN, we use Reverse Channel Attention Module (RCAM) and Reverse Spatial Attention Module (RSAM) to highlight the shallow details of the channel and spatial dimensions. And RCAM and RSAM are used to achieve effective fusion of deep local features with shallow global features to ensure the continuity and integrity of the segmented vessels. In the experimental results for the DRIVE, STARE and CHASE datasets, the evaluation metrics were 0.9712, 0.9822 and 0.9780 for accuracy (Acc), 0.8788, 0.8874 and 0.8352 for sensitivity (Se), 0.9803, 0.9891 and 0.9890 for specificity (Sp), area under the ROC curve(AUC) was 0.9910, 0.9952 and 0.9904, and the F1-Score was 0.8453, 0.8707 and 0.8185. In comparison with existing retinal image segmentation methods, e.g. UNet, R2UNet, DUNet, HAnet, Sine-Net, FANet, etc., our method in three fundus datasets achieved better vessel segmentation performance and results.