{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,9]],"date-time":"2026-09-09T03:51:47Z","timestamp":1788925907074,"version":"build-2803163510"},"reference-count":25,"publisher":"Wiley","issue":"7","license":[{"start":{"date-parts":[[2024,2,14]],"date-time":"2024-02-14T00:00:00Z","timestamp":1707868800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100004543","name":"China Scholarship Council","doi-asserted-by":"publisher","award":["202104910370"],"award-info":[{"award-number":["202104910370"]}],"id":[{"id":"10.13039\/501100004543","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100008658","name":"Deutsches Krebsforschungszentrum","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100008658","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["aapm.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Medical Physics"],"published-print":{"date-parts":[[2024,7]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Background<\/jats:title><jats:p>Digital subtraction angiography (DSA) is a fluoroscopy method primarily used for the diagnosis of cardiovascular diseases (CVDs). Deep learning\u2010based DSA (DDSA) is developed to extract DSA\u2010like images directly from fluoroscopic images, which helps in saving dose while improving image quality. It can also be applied where C\u2010arm or patient motion is present and conventional DSA cannot be applied. However, due to the lack of clinical training data and unavoidable artifacts in DSA targets, current DDSA models still cannot satisfactorily display specific structures, nor can they predict noise\u2010free\u00a0images.<\/jats:p><\/jats:sec><jats:sec><jats:title>Purpose<\/jats:title><jats:p>In this study, we propose a strategy for producing abundant synthetic DSA image pairs in which synthetic DSA targets are free of typical artifacts and noise commonly found in conventional DSA targets for DDSA model\u00a0training.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>More than 7,000 forward\u2010projected computed tomography (CT) images and more than 25,000 synthetic vascular projection images were employed to create contrast\u2010enhanced fluoroscopic images and corresponding DSA images, which were utilized as DSA image pairs for training of the DDSA networks. The CT\u00a0projection images and vascular projection images were generated from eight whole\u2010body CT scans and 1,584 3D vascular skeletons, respectively. All vessel skeletons were generated with stochastic Lindenmayer systems. We trained DDSA models on this synthetic dataset and compared them to the trainings on a clinical DSA dataset, which contains nearly 4,000 fluoroscopic x\u2010ray images obtained from different models of\u00a0C\u2010arms.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>We evaluated DDSA models on clinical fluoroscopic data of different anatomies, including the leg, abdomen, and heart. The results on leg data showed for different methods that training on synthetic data performed similarly and sometimes outperformed training on clinical data. The results on abdomen and cardiac data demonstrated that models trained on synthetic data were able to extract clearer DSA\u2010like images than conventional DSA and models trained on clinical data. The models trained on synthetic data consistently outperformed their clinical data counterparts, achieving higher scores in the quantitative evaluation of peak signal\u2010to\u2010noise ratio (PSNR) and structural similarity index measure (SSIM) metrics for DDSA images, as well as accuracy, precision, and Dice scores for segmentation of the DDSA\u00a0images.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>We proposed an approach to train DDSA networks with synthetic DSA image pairs and extract DSA\u2010like images from contrast\u2010enhanced x\u2010ray images directly. This is a potential tool to aid in\u00a0diagnosis.<\/jats:p><\/jats:sec>","DOI":"10.1002\/mp.16973","type":"journal-article","created":{"date-parts":[[2024,2,14]],"date-time":"2024-02-14T14:54:22Z","timestamp":1707922462000},"page":"4793-4810","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":9,"title":["Training of a deep learning based digital subtraction angiography method using synthetic data"],"prefix":"10.1002","volume":"51","author":[{"given":"Lizhen","family":"Duan","sequence":"first","affiliation":[{"name":"Division of X\u2010Ray Imaging and Computed Tomography German Cancer Research Center (DKFZ) Heidelberg Germany"},{"name":"School of Electronic Electrical and Communication Engineering University of Chinese Academy of Sciences (UCAS) Beijing China"},{"name":"Key Laboratory of Optical Engineering Institute of Optics and Electronics Chinese Academy of Sciences Chengdu China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Elias","family":"Eulig","sequence":"additional","affiliation":[{"name":"Division of X\u2010Ray Imaging and Computed Tomography German Cancer Research Center (DKFZ) Heidelberg Germany"},{"name":"Faculty of Physics and Astronomy Heidelberg University Heidelberg Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Knaup","sequence":"additional","affiliation":[{"name":"Division of X\u2010Ray Imaging and Computed Tomography German Cancer Research Center (DKFZ) Heidelberg Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ralf","family":"Adamus","sequence":"additional","affiliation":[{"name":"Department of Radiology, Neuroradiology and Nuclear Medicine Klinikum N\u00fcrnberg Paracelsus Medical University N\u00fcrnberg Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Lell","sequence":"additional","affiliation":[{"name":"Department of Radiology, Neuroradiology and Nuclear Medicine Klinikum N\u00fcrnberg Paracelsus Medical University N\u00fcrnberg Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marc","family":"Kachelrie\u00df","sequence":"additional","affiliation":[{"name":"Division of X\u2010Ray Imaging and Computed Tomography German Cancer Research Center (DKFZ) Heidelberg Germany"},{"name":"Medical Faculty Heidelberg Heidelberg University Heidelberg Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2024,2,14]]},"reference":[{"key":"e_1_2_9_2_1","doi-asserted-by":"publisher","DOI":"10.1148\/radiology.139.2.7012921"},{"key":"e_1_2_9_3_1","doi-asserted-by":"publisher","DOI":"10.1109\/TNS.1982.4336336"},{"key":"e_1_2_9_4_1","doi-asserted-by":"publisher","DOI":"10.1148\/79.4.658"},{"key":"e_1_2_9_5_1","doi-asserted-by":"publisher","DOI":"10.15836\/ccar.2013.373"},{"key":"e_1_2_9_6_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.proeng.2010.11.070"},{"key":"e_1_2_9_7_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2003.819283"},{"key":"e_1_2_9_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.compmedimag.2013.04.006"},{"key":"e_1_2_9_9_1","doi-asserted-by":"crossref","unstructured":"RonnebergerO FischerP BroxT.U\u2010net: Convolutional networks for biomedical image segmentation. In:International Conference on Medical Image Computing and Computer\u2010assisted Intervention.Springer;2015:234\u2010241.","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"e_1_2_9_10_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2017.07.005"},{"key":"e_1_2_9_11_1","doi-asserted-by":"crossref","unstructured":"EuligE MaierJ KnaupM KoenigT H\u00f6rndlerK Kachelrie\u00dfM.Deep DSA (DDSA): learning mask\u2010free digital subtraction angiography for static and dynamic acquisitions protocols using a deep convolutional neural network. In:ECR 2019 Book of Abstracts Insights into Imaging 10 (Suppl. 1).2019:S379.","DOI":"10.1186\/s13244-019-0713-y"},{"key":"e_1_2_9_12_1","doi-asserted-by":"publisher","DOI":"10.1007\/s11548-019-02040-x"},{"key":"e_1_2_9_13_1","doi-asserted-by":"crossref","unstructured":"EuligE MaierJ KnaupM KoenigT H\u00f6rndlerK Kachelrie\u00dfM.Learned digital subtraction angiography (Deep DSA): method and application to lower extremities. In:15th International Meeting on Fully Three\u2010Dimensional Image Reconstruction in Radiology and Nuclear Medicine. Vol11072.SPIE;2019:360\u2010363.","DOI":"10.1117\/12.2534740"},{"key":"e_1_2_9_14_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jvir.2022.03.010"},{"key":"e_1_2_9_15_1","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.2021203692"},{"key":"e_1_2_9_16_1","doi-asserted-by":"crossref","unstructured":"VepaA ChoiA NakhaeiN et\u00a0al.Weakly\u2010supervised convolutional neural networks for vessel segmentation in cerebral angiography. In:Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision.2022:585\u2010594.","DOI":"10.1109\/WACV51458.2022.00328"},{"key":"e_1_2_9_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2003.819861"},{"key":"e_1_2_9_18_1","doi-asserted-by":"crossref","unstructured":"Galarreta\u2010ValverdeMA MacedoMM MekkaouiC JackowskiMP.Three\u2010dimensional synthetic blood vessel generation using stochastic L\u2010systems. In:Medical Imaging 2013: Image Processing. Vol8669.SPIE;2013:414\u2010419.","DOI":"10.1117\/12.2007532"},{"key":"e_1_2_9_19_1","doi-asserted-by":"publisher","DOI":"10.1147\/sj.41.0025"},{"key":"e_1_2_9_20_1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2021.3130191"},{"key":"e_1_2_9_21_1","unstructured":"ArjovskyM ChintalaS BottouL.Wasserstein generative adversarial networks. In:International Conference on Machine Learning.PMLR;2017:214\u2010223."},{"key":"e_1_2_9_22_1","article-title":"Improved training of Wasserstein Gans","volume":"30","author":"Gulrajani I","year":"2017","journal-title":"Adv Neural Inf Process Syst"},{"key":"e_1_2_9_23_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2022.3192679"},{"key":"e_1_2_9_24_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2022.3177626"},{"key":"e_1_2_9_25_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3250241"},{"key":"e_1_2_9_26_1","unstructured":"KingmaDP BaJ.Adam: A Method for Stochastic Optimization. In:International Conference on Learning Representations ICLR;2015:1\u201013."}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.16973","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,7,9]],"date-time":"2024-07-09T15:53:46Z","timestamp":1720540426000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.16973"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,2,14]]},"references-count":25,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2024,7]]}},"alternative-id":["10.1002\/mp.16973"],"URL":"https:\/\/doi.org\/10.1002\/mp.16973","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,2,14]]},"assertion":[{"value":"2023-06-20","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-01-04","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-02-14","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}