{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,8]],"date-time":"2026-07-08T01:37:50Z","timestamp":1783474670279,"version":"3.55.0"},"reference-count":45,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,10,3]],"date-time":"2024-10-03T00:00:00Z","timestamp":1727913600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Cell Dev. Biol."],"abstract":"<jats:sec><jats:title>Background<\/jats:title><jats:p>Fundus vessel segmentation is vital for diagnosing ophthalmic diseases like central serous chorioretinopathy (CSC), diabetic retinopathy, and glaucoma. Accurate segmentation provides crucial vessel morphology details, aiding the early detection and intervention of ophthalmic diseases. However, current algorithms struggle with fine vessel segmentation and maintaining sensitivity in complex regions. Challenges also stem from imaging variability and poor generalization across multimodal datasets, highlighting the need for more advanced algorithms in clinical practice.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>This paper aims to explore a new vessel segmentation method to alleviate the above problems. We propose a fundus vessel segmentation model based on a combination of double skip connections, deep supervision, and TransUNet, namely DS2TUNet. Initially, the original fundus images are improved through grayscale conversion, normalization, histogram equalization, gamma correction, and other preprocessing techniques. Subsequently, by utilizing the U-Net architecture, the preprocessed fundus images are segmented to obtain the final vessel information. Specifically, the encoder firstly incorporates the ResNetV1 downsampling, dilated convolution downsampling, and Transformer to capture both local and global features, which upgrades its vessel feature extraction ability. Then, the decoder introduces the double skip connections to facilitate upsampling and refine segmentation outcomes. Finally, the deep supervision module introduces multiple upsampling vessel features from the decoder into the loss function, so that the model can learn vessel feature representations more effectively and alleviate gradient vanishing during the training phase.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Extensive experiments on publicly available multimodal fundus datasets such as DRIVE, CHASE_DB1, and ROSE-1 demonstrate that the DS2TUNet model attains F1-scores of 0.8195, 0.8362, and 0.8425, with Accuracy of 0.9664, 0.9741, and 0.9557, Sensitivity of 0.8071, 0.8101, and 0.8586, and Specificity of 0.9823, 0.9869, and 0.9713, respectively. Additionally, the model also exhibits excellent test performance on the clinical fundus dataset CSC, with F1-score of 0.7757, Accuracy of 0.9688, Sensitivity of 0.8141, and Specificity of 0.9801 based on the weight trained on the CHASE_DB1 dataset. These results comprehensively validate that the proposed method obtains good performance in fundus vessel segmentation, thereby aiding clinicians in the further diagnosis and treatment of fundus diseases in terms of effectiveness and feasibility.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fcell.2024.1477819","type":"journal-article","created":{"date-parts":[[2024,10,3]],"date-time":"2024-10-03T04:38:28Z","timestamp":1727930308000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["A fundus vessel segmentation method based on double skip connections combined with deep supervision"],"prefix":"10.3389","volume":"12","author":[{"given":"Qingyou","family":"Liu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fen","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jianxin","family":"Shen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jianguo","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cheng","family":"Wan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiangzhong","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhipeng","family":"Yan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jin","family":"Yao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,10,3]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"014006","DOI":"10.1117\/1.JMI.6.1.014006","article-title":"Recurrent residual U-Net for medical image segmentation","volume":"6","author":"Alom","year":"2019","journal-title":"J. Med. Imaging"},{"key":"B2","doi-asserted-by":"publisher","first-page":"118313","DOI":"10.1016\/j.eswa.2022.118313","article-title":"Width attention based convolutional neural network for retinal vessel segmentation","volume":"209","author":"Alvarado-Carrillo","year":"2022","journal-title":"Expert Syst. Appl."},{"key":"B3","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1016\/j.media.2014.08.002","article-title":"Trainable COSFIRE filters for vessel delineation with application to retinal images","volume":"19","author":"Azzopardi","year":"2015","journal-title":"Med. Image Anal."},{"key":"B4","first-page":"205","article-title":"Swin-UNet: UNet-like pure transformer for medical image segmentation","volume-title":"Computer vision - ECCV 2022 workshopsLecture notes in computer science","author":"Cao","year":"2023"},{"key":"B5","first-page":"213","article-title":"End-to-end object detection with transformers","author":"Carion","year":"2020"},{"key":"B6","doi-asserted-by":"publisher","first-page":"263","DOI":"10.1109\/42.34715","article-title":"Detection of blood vessels in retinal images using two-dimensional matched filters","volume":"8","author":"Chaudhuri","year":"1989","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2102.04306","article-title":"TransUNet: transformers make strong encoders for medical image segmentation","author":"Chen","year":"2021","journal-title":"arXiv Prepr. arXiv:2102.04306"},{"key":"B8","doi-asserted-by":"publisher","first-page":"161","DOI":"10.1186\/s13195-020-00724-0","article-title":"Retinal microvasculature dysfunction is associated with Alzheimer's disease and mild cognitive impairment","volume":"12","author":"Chua","year":"2020","journal-title":"Alz Res. Ther."},{"key":"B9","article-title":"An image is worth 16x16 words: transformers for image recognition at scale","author":"Dosovitskiy","year":"2021"},{"key":"B10","doi-asserted-by":"publisher","first-page":"175013","DOI":"10.1088\/1361-6560\/ac1c4c","article-title":"UNet retinal blood vessel segmentation algorithm based on improved pyramid pooling method and attention mechanism","volume":"66","author":"Du","year":"2021","journal-title":"Phys. Med. and Biol."},{"key":"B11","doi-asserted-by":"publisher","first-page":"268","DOI":"10.1016\/j.neucom.2018.10.098","article-title":"CcNet: a cross-connected convolutional network for segmenting retinal vessels using multi-scale features","volume":"392","author":"Feng","year":"2020","journal-title":"Neurocomputing"},{"key":"B12","doi-asserted-by":"publisher","first-page":"407","DOI":"10.1016\/j.cmpb.2012.03.009","article-title":"Blood vessel segmentation methodologies in retinal images\u2014a survey","volume":"108","author":"Fraz","year":"2012","journal-title":"Comput. Methods Programs Biomed."},{"key":"B13","doi-asserted-by":"publisher","first-page":"132","DOI":"10.1007\/978-3-319-46723-8_16","article-title":"Deepvessel: retinal vessel segmentation via deep learning and conditional random field","volume":"2016","author":"Fu","year":"2016","journal-title":"Med. Image Comput. Computer-Assisted Intervention\u2013MICCAI"},{"key":"B14","doi-asserted-by":"publisher","first-page":"2281","DOI":"10.1109\/TMI.2019.2903562","article-title":"CE-Net: context encoder network for 2D medical image segmentation","volume":"38","author":"Gu","year":"2019","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B15","doi-asserted-by":"publisher","first-page":"5689","DOI":"10.1007\/s12652-021-03422-3","article-title":"DPN: detail-preserving network with high resolution representation for efficient segmentation of retinal vessels","volume":"14","author":"Guo","year":"2023","journal-title":"J. Ambient Intell. Humaniz. Comput."},{"key":"B16","first-page":"2441","article-title":"UCTransnet: rethinking the skip connections in U-Net from a channel-wise perspective with transformer","author":"Haonan","year":"2022"},{"key":"B17","first-page":"770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B18","doi-asserted-by":"publisher","first-page":"149","DOI":"10.1016\/j.knosys.2019.04.025","article-title":"DUNet: a deformable network for retinal vessel segmentation","volume":"178","author":"Jin","year":"","journal-title":"Knowledge-Based Syst."},{"key":"B19","doi-asserted-by":"publisher","first-page":"149","DOI":"10.1016\/j.knosys.2019.04.025","article-title":"DUNet: a deformable network for retinal vessel segmentation","volume":"178","author":"Jin","year":"","journal-title":"Knowledge-Based Syst."},{"key":"B20","doi-asserted-by":"publisher","first-page":"43207","DOI":"10.1007\/s11042-023-15433-7","article-title":"IterMiUnet: a lightweight architecture for automatic blood vessel segmentation","volume":"82","author":"Kumar","year":"2023","journal-title":"Multimedia Tools Appl."},{"key":"B21","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1155\/2021\/5595026","article-title":"Medicalguard: U-net model robust against adversarially perturbed images","volume":"2021","author":"Kwon","year":"2021","journal-title":"Secur. Commun. Netw."},{"key":"B22","first-page":"1958","article-title":"MAU-Net: a retinal vessels segmentation method","author":"Li","year":"2020"},{"key":"B23","doi-asserted-by":"publisher","first-page":"7600","DOI":"10.1016\/j.eswa.2011.12.046","article-title":"Vessel segmentation and width estimation in retinal images using multiscale production of matched filter responses","volume":"39","author":"Li","year":"2012","journal-title":"Expert Syst. Appl."},{"key":"B24","doi-asserted-by":"publisher","first-page":"109","DOI":"10.1109\/TMI.2015.2457891","article-title":"A cross-modality learning approach for vessel segmentation in retinal images","volume":"35","author":"Li","year":"2016","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B25","doi-asserted-by":"publisher","first-page":"104087","DOI":"10.1016\/j.bspc.2022.104087","article-title":"ResDO-UNet: a deep residual network for accurate retinal vessel segmentation from fundus images","volume":"79","author":"Liu","year":"2023","journal-title":"Biomed. Signal Process. Control"},{"key":"B26","doi-asserted-by":"publisher","first-page":"32826","DOI":"10.1109\/access.2020.2974027","article-title":"Attention guided U-Net with atrous convolution for accurate retinal vessels segmentation","volume":"8","author":"Lv","year":"2020","journal-title":"IEEE Access"},{"key":"B27","doi-asserted-by":"publisher","first-page":"928","DOI":"10.1109\/TMI.2020.3042802","article-title":"ROSE: a retinal OCT-angiography vessel segmentation dataset and new model","volume":"40","author":"Ma","year":"2020","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B28","first-page":"721","article-title":"CS-Net: channel and spatial attention network for curvilinear structure segmentation","volume-title":"Medical image computing and computer-assisted intervention \u2013 miccai 2019","author":"Mou","year":"2019"},{"key":"B29","doi-asserted-by":"publisher","first-page":"101874","DOI":"10.1016\/j.media.2020.101874","article-title":"CS2-Net: deep learning segmentation of curvilinear structures in medical imaging","volume":"67","author":"Mou","year":"2021","journal-title":"Med. Image Anal."},{"key":"B30","doi-asserted-by":"publisher","first-page":"16","DOI":"10.1109\/TBME.2016.2535311","article-title":"A discriminatively trained fully connected conditional random field model for blood vessel segmentation in fundus images","volume":"64","author":"Orlando","year":"2016","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B31","first-page":"234","article-title":"U-Net: convolutional networks for biomedical image segmentation","volume-title":"Medical image computing and computer-assisted intervention (MICCAI)","author":"Ronneberger","year":"2015"},{"key":"B32","doi-asserted-by":"publisher","first-page":"197","DOI":"10.1016\/j.media.2019.01.012","article-title":"Attention gated networks: learning to leverage salient regions in medical images","volume":"53","author":"Schlemper","year":"2019","journal-title":"Med. Image Anal."},{"key":"B33","doi-asserted-by":"publisher","first-page":"640","DOI":"10.1109\/TPAMI.2016.2572683","article-title":"Fully convolutional networks for semantic segmentation","volume":"39","author":"Shelhamer","year":"2017","journal-title":"IEEE Trans. Pattern Analysis Mach. Intell."},{"key":"B34","doi-asserted-by":"publisher","first-page":"1196191","DOI":"10.3389\/fcell.2023.1196191","article-title":"Attention-guided cascaded network with pixel-importance-balance loss for retinal vessel segmentation","volume":"11","author":"Su","year":"2023","journal-title":"Front. Cell. Dev. Biol."},{"key":"B35","first-page":"3660","article-title":"The fully convolutional transformer for medical image segmentation","author":"Tragakis","year":"2023"},{"key":"B36","first-page":"6000","article-title":"Attention is all you need","author":"Vaswani","year":"2017"},{"key":"B37","doi-asserted-by":"publisher","first-page":"035008","DOI":"10.1088\/1361-6560\/acb19a","article-title":"Cross-convolutional transformer for automated multi-organs segmentation in a variety of medical images","volume":"68","author":"Wang","year":"2023","journal-title":"Phys. Med. Biol."},{"key":"B38","first-page":"548","article-title":"Pyramid vision transformer: a versatile backbone for dense prediction without convolutions","author":"Wang","year":"2022"},{"key":"B39","first-page":"264","article-title":"Vessel-Net: retinal vessel segmentation under multi-path supervision","author":"Wu","year":"2019"},{"key":"B40","doi-asserted-by":"publisher","first-page":"1912","DOI":"10.1109\/TBME.2018.2828137","article-title":"Joint segment-level and pixel-wise losses for deep learning based retinal vessel segmentation","volume":"65","author":"Yan","year":"","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B41","doi-asserted-by":"publisher","first-page":"1427","DOI":"10.1109\/JBHI.2018.2872813","article-title":"A three-stage deep learning model for accurate retinal vessel segmentation","volume":"23","author":"Yan","year":"","journal-title":"IEEE J. Biomed. Health Inf."},{"key":"B42","first-page":"553","article-title":"TransNUNet: using attention mechanism for whole heart segmentation","author":"Yang","year":"2022"},{"key":"B43","doi-asserted-by":"publisher","first-page":"1181305","DOI":"10.3389\/fcell.2023.1181305","article-title":"Joint conditional generative adversarial networks for eyelash artifact removal in ultra-wide-field fundus images","volume":"11","author":"Zhang","year":"2023","journal-title":"Front. Cell. Dev. Biol."},{"key":"B44","doi-asserted-by":"publisher","first-page":"749","DOI":"10.1109\/lgrs.2018.2802944","article-title":"Road extraction by deep residual u-net","volume":"15","author":"Zhang","year":"2018","journal-title":"IEEE Geoscience Remote Sens. Lett."},{"key":"B45","first-page":"6877","article-title":"Rethinking semantic segmentation from a sequence-to-sequence perspective with transformers","author":"Zheng","year":"2021"}],"container-title":["Frontiers in Cell and Developmental Biology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fcell.2024.1477819\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,3]],"date-time":"2024-10-03T04:38:30Z","timestamp":1727930310000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fcell.2024.1477819\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,10,3]]},"references-count":45,"alternative-id":["10.3389\/fcell.2024.1477819"],"URL":"https:\/\/doi.org\/10.3389\/fcell.2024.1477819","relation":{},"ISSN":["2296-634X"],"issn-type":[{"value":"2296-634X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,10,3]]},"article-number":"1477819"}}