Fundus Image Based Retinal Vessel Segmentation Utilizing a Fast and Accurate Fully Convolutional Network

Lyu, Junyan; Cheng, Pujin; Tang, Xiaoying

doi:10.1007/978-3-030-32956-3_14

Cited by 16 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It shows that the proposed BEFD-UNet ranks first in four metrics except for F1 score. More specifically, it achieves the highest accuracy (1.05% higher than the second best [16]), the optimal AUC (0.41% higher than the suboptimal result [15]) as well as the best sensitivity (0.80% higher than the previous highest score [26]). Additionally, it is worth mentioning that the F1 score (0.8267) of our method is relatively close to the best score (0.8270) [20].…”

Section: Resultsmentioning

confidence: 99%

“…Alom [1] 2019 0.7792 0.9813 0.8171 0.9556 0.9784 Jin [12] 2019 0.7963 0.9800 0.8237 0.9566 0.9802 Mou [16] 2019 0.8126 0.9788 -0.9594 0.9796 Wu [22] 2019 0.8038 0.9802 -0.9578 0.9821 Lyu [15] 2019 0.7940 0.9820 -0.9579 0.9826 Wang [20] 2019 0.7940 0.9816 0.8270 0.9567 0.9772 Zhou [26] 2019 0.8135 0.9768 0.8249 0.9560 0.9739 Li [14] 2020 0.7791 0.9831 0.8218 0.9574 0.9813 UNet 2020 0.7887 0.9861 0.8140 0.9686 0.9836 BEFD-UNet 2020 0.8215 0.9845 0.8267 0.9701 0.9867 making it possible to pay more attention to the object contours. When the feature maps in the encoder path are multiplied (element-wise) by such attention map, the object boundaries are enhanced accordingly.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

BEFD: Boundary Enhancement and Feature Denoising for Vessel Segmentation

Zhang

Zhao

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Blood vessel segmentation is crucial for many diagnostic and research applications. In recent years, CNN-based models have leaded to breakthroughs in the task of segmentation, however, such methods usually lose high-frequency information like object boundaries and subtle structures, which are vital to vessel segmentation. To tackle this issue, we propose Boundary Enhancement and Feature Denoising (BEFD) module to facilitate the network ability of extracting boundary information in semantic segmentation, which can be integrated into arbitrary encoderdecoder architecture in an end-to-end way. By introducing Sobel edge detector, the network is able to acquire additional edge prior, thus enhancing boundary in an unsupervised manner for medical image segmentation. In addition, we also utilize a denoising block to reduce the noise hidden in the low-level features. Experimental results on retinal vessel dataset and angiocarpy dataset demonstrate the superior performance of the new BEFD module.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

BEFD: Boundary Enhancement and Feature Denoising for Vessel Segmentation

Zhang

Zhao

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…DRIU [27] is the first method using CNN to understand fundus images, solving both retinal vessel segmentation and OD segmentation. Lyu et al [5] proposed a novel fully convolutional network utilizing separable spatial and channel flow and densely adjacent vessel prediction to better capture spatial correlations between vessels. Fu et al [28] performed joint OD/OC segmentation on polar coordinates using a multi-scale U-Net, which balanced the area proportion of OD and OC.…”

Section: A Fundus Image Segmentationmentioning

confidence: 99%

AADG: Automatic Augmentation for Domain Generalization on Retinal Image Segmentation

Lyu

Zhang

Huang

et al. 2022

IEEE Trans. Med. Imaging

Self Cite

View full text Add to dashboard Cite

Convolutional neural networks have been widely applied to medical image segmentation and have achieved considerable performance. However, the performance may be significantly affected by the domain gap between training data (source domain) and testing data (target domain). To address this issue, we propose a data manipulation based domain generalization method, called Automated Augmentation for Domain Generalization (AADG). Our AADG framework can effectively sample data augmentation policies that generate novel domains and diversify the training set from an appropriate search space. Specifically, we introduce a novel proxy task maximizing the diversity among multiple augmented novel domains as measured by the Sinkhorn distance in a unit sphere space, making automated augmentation tractable. Adversarial training and deep reinforcement learning are employed to efficiently search the objectives. Quantitative and qualitative experiments on 11 publicly-accessible fundus image datasets (four for retinal vessel segmentation, four for optic disc and cup (OD/OC) segmentation and three for retinal lesion segmentation) are comprehensively performed. Two OCTA datasets for retinal vasculature segmentation are further involved to validate cross-modality generalization. Our proposed AADG exhibits state-of-theart generalization performance and outperforms existing approaches by considerable margins on retinal vessel, OD/OC and lesion segmentation tasks. The learned policies are empirically validated to be model-agnostic and can transfer well to other models. The source code is available at https://github.com/CRazorback/AADG.

show abstract

“…With the capability of highly representative feature extraction, convolutional neural networks (CNNs) have been proposed to tackle different tasks. They have also been widely used in the medical image analysis realm [8][9][10][11][12]. In DR grading, ref.…”

Section: Introductionmentioning

confidence: 99%

Identifying the Key Components in ResNet-50 for Diabetic Retinopathy Grading from Fundus Images: A Systematic Investigation

et al. 2023

Self Cite

View full text Add to dashboard Cite

Although deep learning-based diabetic retinopathy (DR) classification methods typically benefit from well-designed architectures of convolutional neural networks, the training setting also has a non-negligible impact on prediction performance. The training setting includes various interdependent components, such as an objective function, a data sampling strategy, and a data augmentation approach. To identify the key components in a standard deep learning framework (ResNet-50) for DR grading, we systematically analyze the impact of several major components. Extensive experiments are conducted on a publicly available dataset EyePACS. We demonstrate that (1) the DR grading framework is sensitive to input resolution, objective function, and composition of data augmentation; (2) using mean square error as the loss function can effectively improve the performance with respect to a task-specific evaluation metric, namely the quadratically weighted Kappa; (3) utilizing eye pairs boosts the performance of DR grading and; (4) using data resampling to address the problem of imbalanced data distribution in EyePACS hurts the performance. Based on these observations and an optimal combination of the investigated components, our framework, without any specialized network design, achieves a state-of-the-art result (0.8631 for Kappa) on the EyePACS test set (a total of 42,670 fundus images) with only image-level labels. We also examine the proposed training practices on other fundus datasets and other network architectures to evaluate their generalizability. Our codes and pre-trained model are available online.

show abstract

Fundus Image Based Retinal Vessel Segmentation Utilizing a Fast and Accurate Fully Convolutional Network

Cited by 16 publications

References 12 publications

BEFD: Boundary Enhancement and Feature Denoising for Vessel Segmentation

BEFD: Boundary Enhancement and Feature Denoising for Vessel Segmentation

AADG: Automatic Augmentation for Domain Generalization on Retinal Image Segmentation

Identifying the Key Components in ResNet-50 for Diabetic Retinopathy Grading from Fundus Images: A Systematic Investigation

Contact Info

Product

Resources

About