{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T00:02:41Z","timestamp":1780617761659,"version":"3.54.1"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1012231","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2024,7,2]],"date-time":"2024-07-02T00:00:00Z","timestamp":1719878400000}}],"reference-count":32,"publisher":"Public Library of Science (PLoS)","issue":"6","license":[{"start":{"date-parts":[[2024,6,20]],"date-time":"2024-06-20T00:00:00Z","timestamp":1718841600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100014013","name":"UK Research and Innovation","doi-asserted-by":"crossref","award":["EP\/S021612\/1, MR\/S032290\/1"],"award-info":[{"award-number":["EP\/S021612\/1, MR\/S032290\/1"]}],"id":[{"id":"10.13039\/100014013","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000274","name":"the British Heart Foundation","doi-asserted-by":"crossref","award":["NH\/18\/1\/33511, PG\/17\/6\/32797"],"award-info":[{"award-number":["NH\/18\/1\/33511, PG\/17\/6\/32797"]}],"id":[{"id":"10.13039\/501100000274","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000327","name":"Heart Research UK","doi-asserted-by":"crossref","award":["RG2661\/17\/20"],"award-info":[{"award-number":["RG2661\/17\/20"]}],"id":[{"id":"10.13039\/501100000327","id-type":"DOI","asserted-by":"crossref"}]},{"name":"European Research Council","award":["ERC-2017-StG-757923"],"award-info":[{"award-number":["ERC-2017-StG-757923"]}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>Computational fluid dynamics (CFD) can be used for non-invasive evaluation of hemodynamics. However, its routine use is limited by labor-intensive manual segmentation, CFD mesh creation, and time-consuming simulation. This study aims to train a deep learning model to both generate patient-specific volume-meshes of the pulmonary artery from 3D cardiac MRI data and directly estimate CFD flow fields. This proof-of-concept study used 135 3D cardiac MRIs from both a public and private dataset. The pulmonary arteries in the MRIs were manually segmented and converted into volume-meshes. CFD simulations were performed on ground truth meshes and interpolated onto point-point correspondent meshes to create the ground truth dataset. The dataset was split 110\/10\/15 for training, validation, and testing. Image2Flow, a hybrid image and graph convolutional neural network, was trained to transform a pulmonary artery template to patient-specific anatomy and CFD values, taking a specific inlet velocity as an additional input. Image2Flow was evaluated in terms of segmentation, and the accuracy of predicted CFD was assessed using node-wise comparisons. In addition, the ability of Image2Flow to respond to increasing inlet velocities was also evaluated. Image2Flow achieved excellent segmentation accuracy with a median Dice score of 0.91 (IQR: 0.86\u20130.92). The median node-wise normalized absolute error for pressure and velocity magnitude was 11.75% (IQR: 9.60\u201315.30%) and 9.90% (IQR: 8.47\u201311.90), respectively. Image2Flow also showed an expected response to increased inlet velocities with increasing pressure and velocity values. This proof-of-concept study has shown that it is possible to simultaneously perform patient-specific volume-mesh based segmentation and pressure and flow field estimation using Image2Flow. Image2Flow completes segmentation and CFD in ~330ms, which is ~5000 times faster than manual methods, making it more feasible in a clinical environment.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1012231","type":"journal-article","created":{"date-parts":[[2024,6,20]],"date-time":"2024-06-20T17:41:30Z","timestamp":1718905290000},"page":"e1012231","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":13,"title":["Image2Flow: A proof-of-concept hybrid image and graph convolutional neural network for rapid patient-specific pulmonary artery segmentation and CFD flow field calculation from 3D cardiac MRI data"],"prefix":"10.1371","volume":"20","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0470-311X","authenticated-orcid":true,"given":"Tina","family":"Yao","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Endrit","family":"Pajaziti","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Quail","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Silvia","family":"Schievano","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9792-2022","authenticated-orcid":true,"given":"Jennifer","family":"Steeden","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4292-6456","authenticated-orcid":true,"given":"Vivek","family":"Muthurangu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2024,6,20]]},"reference":[{"key":"pcbi.1012231.ref001","doi-asserted-by":"crossref","article-title":"Computational modelling for congenital heart disease: how far are we from clinical translation?","author":"G Biglino","DOI":"10.1136\/heartjnl-2016-310423"},{"key":"pcbi.1012231.ref002","doi-asserted-by":"crossref","first-page":"151141","DOI":"10.3389\/fped.2015.00107","article-title":"Using 4D Cardiovascular Magnetic Resonance Imaging to Validate Computational Fluid Dynamics: A Case Study.","volume":"3","author":"G Biglino","year":"2015","journal-title":"Front Pediatr."},{"key":"pcbi.1012231.ref003","doi-asserted-by":"crossref","first-page":"587","DOI":"10.1097\/MOP.0000000000000269","article-title":"Computational modeling and engineering in pediatric and congenital heart disease HHS Public Access","volume":"27","author":"AL Marsden","year":"2015","journal-title":"Curr Opin Pediatr"},{"key":"pcbi.1012231.ref004","doi-asserted-by":"crossref","first-page":"1281","DOI":"10.1016\/j.jcin.2015.06.015","article-title":"Biomechanical Modeling to Improve Coronary Artery Bifurcation Stenting.","volume":"8","author":"AP Antoniadis","year":"2015","journal-title":"JACC Cardiovasc Interv"},{"key":"pcbi.1012231.ref005","first-page":"133","article-title":"Computational Simulations for Aortic Coarctation: Representative Results From a Sampling of Patients","author":"JF LaDisa","year":"2011","journal-title":"J Biomech Eng"},{"key":"pcbi.1012231.ref006","doi-asserted-by":"crossref","first-page":"1605","DOI":"10.1109\/TMI.2011.2135375","article-title":"A statistical model for quantification and prediction of cardiac remodelling: Application to tetralogy of fallot","volume":"30","author":"T Mansi","year":"2011","journal-title":"IEEE Trans Med Imaging"},{"key":"pcbi.1012231.ref007","doi-asserted-by":"crossref","first-page":"111","DOI":"10.1016\/0021-9290(95)00021-6","article-title":"A numerical fluid mechanical study of repaired congenital heart defects. Application to the total cavopulmonary connection","volume":"29","author":"G Dubini","year":"1996","journal-title":"J Biomech"},{"key":"pcbi.1012231.ref008","doi-asserted-by":"crossref","first-page":"e02737","DOI":"10.1002\/cnm.2737","article-title":"Uncertainty quantification in virtual surgery hemodynamics predictions for single ventricle palliation","volume":"32","author":"DE Schiavazzi","year":"2016","journal-title":"Int J Numer Method Biomed Eng"},{"key":"pcbi.1012231.ref009","article-title":"A review of segmentation methods in short axis cardiac MR images","author":"C Petitjean","year":"2010"},{"key":"pcbi.1012231.ref010","author":"K Huang","year":"2021","journal-title":"MACHINE LEARNING-BASED OPTIMAL MESH GENERATION IN COMPUTATIONAL FLUID DYNAMICS A PREPRINT"},{"key":"pcbi.1012231.ref011","doi-asserted-by":"crossref","first-page":"18","DOI":"10.1136\/heartjnl-2015-308044","article-title":"Computational fluid dynamics modelling in cardiovascular medicine","volume":"102","author":"PD Morris","year":"2016","journal-title":"Heart"},{"key":"pcbi.1012231.ref012","article-title":"Computational Fluid Dynamics Applied to Cardiac Computed Tomography for Noninvasive Quantification of Fractional Flow Reserve Scientific Basis.","author":"CA Taylor"},{"key":"pcbi.1012231.ref013","doi-asserted-by":"crossref","DOI":"10.3389\/fcvm.2020.00025","article-title":"Deep Learning for Cardiac Image Segmentation: A Review.","volume":"7","author":"C Chen","year":"2020","journal-title":"Front Cardiovasc Med."},{"key":"pcbi.1012231.ref014","doi-asserted-by":"crossref","first-page":"57","DOI":"10.1186\/s12968-022-00891-z","article-title":"Automatic segmentation of the great arteries for computational hemodynamic assessment","volume":"24","author":"J Montalt-Tordera","year":"2022","journal-title":"Journal of Cardiovascular Magnetic Resonance"},{"key":"pcbi.1012231.ref015","first-page":"299","volume-title":"Medical Image Computing and Computer Assisted Intervention\u2013MICCAI 2020.","author":"U Wickramasinghe","year":"2020"},{"key":"pcbi.1012231.ref016","doi-asserted-by":"crossref","first-page":"102222","DOI":"10.1016\/j.media.2021.102222","article-title":"A deep-learning approach for direct whole-heart mesh reconstruction","volume":"74","author":"F Kong","year":"2021","journal-title":"Med Image Anal"},{"key":"pcbi.1012231.ref017","doi-asserted-by":"crossref","first-page":"533","DOI":"10.1109\/TMI.2022.3219284","article-title":"Learning Whole Heart Mesh Generation From Patient Images for Computational Simulations","volume":"42","author":"F Kong","year":"2023","journal-title":"IEEE Trans Med Imaging"},{"key":"pcbi.1012231.ref018","first-page":"550","volume-title":"Medical Image Computing and Computer Assisted Intervention\u2013MICCAI 2021.","author":"F Kong","year":"2021"},{"key":"pcbi.1012231.ref019","doi-asserted-by":"crossref","first-page":"e1011055","DOI":"10.1371\/journal.pcbi.1011055","article-title":"Shape-driven deep neural networks for fast acquisition of aortic 3D pressure and velocity flow fields.","volume":"19","author":"E Pajaziti","year":"2023","journal-title":"PLoS Comput Biol"},{"key":"pcbi.1012231.ref020","doi-asserted-by":"crossref","first-page":"1519","DOI":"10.1007\/s10237-017-0903-9","article-title":"A machine learning approach to investigate the relationship between shape features and numerically predicted risk of ascending aortic aneurysm.","volume":"16","author":"L Liang","year":"2017","journal-title":"Biomech Model Mechanobiol"},{"key":"pcbi.1012231.ref021","doi-asserted-by":"crossref","first-page":"109544","DOI":"10.1016\/j.jbiomech.2019.109544","article-title":"A feasibility study of deep learning for predicting hemodynamics of human thoracic aorta","volume":"99","author":"L Liang","year":"2020","journal-title":"J Biomech"},{"key":"pcbi.1012231.ref022","doi-asserted-by":"crossref","first-page":"1221541","DOI":"10.3389\/fcvm.2023.1221541","article-title":"A physics-based machine learning technique rapidly reconstructs the wall-shear stress and pressure fields in coronary arteries.","volume":"10","author":"B Morgan","year":"2023","journal-title":"Front Cardiovasc Med"},{"key":"pcbi.1012231.ref023","first-page":"1","article-title":"Prediction of 3D Cardiovascular hemodynamics before and after coronary artery bypass surgery via deep learning","volume":"4","author":"G Li","year":"2021","journal-title":"Commun Biol"},{"key":"pcbi.1012231.ref024","doi-asserted-by":"crossref","first-page":"525","DOI":"10.1093\/ehjdh\/ztac058","article-title":"An artificial intelligence-based platform for automatically estimating time-averaged wall shear stress in the ascending aorta","volume":"3","author":"L Lv","year":"2022","journal-title":"European Heart Journal\u2014Digital Health"},{"key":"pcbi.1012231.ref025","doi-asserted-by":"crossref","first-page":"107676","DOI":"10.1016\/j.compbiomed.2023.107676","article-title":"Learning reduced-order models for cardiovascular simulations with graph neural networks","volume":"168","author":"L Pegolotti","year":"2024","journal-title":"Comput Biol Med"},{"key":"pcbi.1012231.ref026","author":"D Grattarola","year":"2020","journal-title":"Graph Neural Networks in TensorFlow and Keras with Spektral"},{"key":"pcbi.1012231.ref027","doi-asserted-by":"crossref","first-page":"101537","DOI":"10.1016\/j.media.2019.101537","article-title":"Evaluation of algorithms for Multi-Modality Whole Heart Segmentation: An open-access grand challenge","volume":"58","author":"X Zhuang","year":"2019","journal-title":"Med Image Anal"},{"key":"pcbi.1012231.ref028","doi-asserted-by":"crossref","first-page":"745","DOI":"10.21105\/joss.00745","article-title":"The Vascular Modeling Toolkit: A Python Library for the Analysis of Tubular Structures in Medical Images.","volume":"3","author":"R Izzo","year":"2018","journal-title":"J Open Source Softw."},{"key":"pcbi.1012231.ref029","article-title":"Measurement of instantaneous blood flow velocity and pressure in conscious man with a catheter-tip velocity probe","volume":"XL","author":"IT Gabe","year":"1969","journal-title":"Circulation"},{"key":"pcbi.1012231.ref030","author":"F Kong","year":"2023","journal-title":"SDF4CHD: Generative Modeling of Cardiac Anatomies with Congenital Heart Defects."},{"key":"pcbi.1012231.ref031","doi-asserted-by":"crossref","first-page":"102","DOI":"10.1016\/j.jbiomech.2012.10.012","article-title":"Inflow boundary conditions for image-based computational hemodynamics: Impact of idealized versus measured velocity profiles in the human aorta","volume":"46","author":"U Morbiducci","year":"2013","journal-title":"J Biomech"},{"key":"pcbi.1012231.ref032","doi-asserted-by":"crossref","DOI":"10.1115\/1.4037857","article-title":"Impact of patient-specific inflow velocity profile on hemodynamics of the thoracic aorta","volume":"140","author":"P Youssefi","year":"2018","journal-title":"J Biomech Eng"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1012231","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2024,7,2]],"date-time":"2024-07-02T00:00:00Z","timestamp":1719878400000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1012231","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,7,2]],"date-time":"2024-07-02T19:31:23Z","timestamp":1719948683000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1012231"}},"subtitle":[],"editor":[{"given":"Alison","family":"Marsden","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2024,6,20]]},"references-count":32,"journal-issue":{"issue":"6","published-online":{"date-parts":[[2024,6,20]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1012231","relation":{"new_version":[{"id-type":"doi","id":"10.1371\/journal.pcbi.1012231","asserted-by":"object"}]},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,6,20]]}}}