AUTHOR=Hu Jingfei , Wang Hua , Cao Zhaohui , Wu Guang , Jonas Jost B. , Wang Ya Xing , Zhang Jicong 

TITLE=Automatic Artery/Vein Classification Using a Vessel-Constraint Network for Multicenter Fundus Images

JOURNAL=Frontiers in Cell and Developmental Biology

VOLUME=Volume 9 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/cell-and-developmental-biology/articles/10.3389/fcell.2021.659941

DOI=10.3389/fcell.2021.659941

ISSN=2296-634X

ABSTRACT=Retinal blood vessel morphological abnormalities are generally associated with cardiovascular, cerebrovascular, and systemic diseases, automatic artery/vein (A/V) classification is particularly important for medical image analysis and clinical decision making. However, current method still has some limitations in A/V classification, especially the blood vessel edge and end error problems caused by the single scale and the blurred boundary of the artery/vein. To alleviate these problems, in this work we propose a vessel-constraint network (VC-Net) that utilize information of vessel distribution and edge to enhance A/V classification, which is a high precision A/V classification model based on data fusion. Especially, VC-Net introduces a vessel-constraint (VC) module that combines the local and global vessel information to generate a weight map to constraint the A/V features, which suppressing the background-prone features and enhancing the edge and end features of blood vessels. In addition, the VC-Net employees a multi-scale feature (MSF) module to extract blood vessel information with different scales to improve the feature extraction capability and robustness of the model. And the VC-Net can get vessel segmentation result simultaneously. The proposed method is tested on publicly available fundus image datasets with different scales, namely DRIVE, LES, HRF, and validated on two newly-created multi-center datasets: Tongren and Kailuan. We achieve the balance accuracy of 0.9554, F1-scores of 0.7616, and 0.7971 for the arteries and veins on DRIVE dataset. The experimental results prove that proposed model achieves competitive performance in A/V classification and vessel segmentation tasks compared with state-of-the-art methods. Finally, we test the Kailuan dataset with other trained fusion datasets, the results also show good robustness. To promote research in this area, the Tongren dataset and source code will be publicly available.