Generating Short-Axis DENSE Images from 4D XCAT Phantoms: A Proof-of-Concept Study

TY - CHAP

T1 - Generating Short-Axis DENSE Images from 4D XCAT Phantoms: A Proof-of-Concept Study

AU - Barbaroux, Hugo

AU - Loecher, Michael

AU - Kunze, Karl P.

AU - Neji, Radhouene

AU - Ennis, Daniel B.

AU - Nielles-Vallespin, Sonia

AU - Scott, Andrew D.

AU - Young, Alistair A.

N1 - Funding Information: Acknowledegments. This work was supported by the British Heart Foundation Centre of Research Excellence at Imperial College London (RE/18/4/34215). K. Vimalesvaran and S. Zaman were supported by UKRI Centre for Doctoral Training in AI for Healthcare grant number EP/S023283/1. The project was partly supported by a Rosetrees Interdisciplinary Award. Funding Information: Acknowledgements. The study was approved by the Ethics Committee of the University Medical Center Ljubljana, Slovenia, under 0120-133/2021/3 and 0120-312/2022/3, and supported by the Slovenian Research Agency (ARRS) under grants J2-4453 and P2-0232, and by the University Medical Center Ljubljana, Slovenia, under grant 20190174. Funding Information: Acknowledgements. This work has been supported by the French government through the National Research Agency (ANR) Investments in the Future with 3IA Côte d’Azur (ANR-19-P3IA-0002) and by Inria PhD funding. The authors are grateful to the OPAL infrastructure from Université Côte d’Azur for providing resources and support. Funding Information: Acknowledgements. The authors thank the Stanford Research Computing Center for computational resources (Sherlock HPC Cluster). This work is funded by a National Science Foundation Graduate Research Fellowship (DGE-1656518) to PJN and by the Stanford Maternal and Child Health Research Institute (award K99HL161313) to MRP. Funding Information: gen, Germany, for providing the 4D flow work-in-progress package. Malak Sabry is supported by a PhD educational grant from Siemens Healthineers and the Magdi Yacoub Foundation. Funding Information: Acknowledgements. This work is funded by a EPSRC grant (EP/X023826/1). The study was also supported by the Wellcome/EPSRC Centre for Medical Engineering (WT203148/Z/16/Z) and the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. Funding Information: Acknowledgements. This work has been performed under FI 2022 grant number 00237, awarded by the Agency for Management of University and Research Grants (AGAUR), Generalitat de Catalunya. It has been partially funded by European Union-NextGenerationEU, Ministry of Universities and Recovery, Transformation and Resilience Plan, through a call from Pompeu Fabra University (Barcelona). Funding Information: Acknowledgments. This Project has received funding from the European Unions Horizon research and innovation programme under the Marie Skodowska-Curie grant agreement No. 860974 and by the French National Research Agency, grant references ANR-10-IAHU04-LIRYC and ANR-11-EQPX-0030. The study was carried out as part of the PersonalizeAF project in collaboration with EP-Solutions SA. Funding Information: Youth and Sports of the Czech Republic under the OP RDE grants number CZ2.11/0/0/16 019/0000765 and by the Ministry of Health of the Czech Republic project No. NV19-08-00071. This work was also supported by the Inria-UTSW Associated Team TOFMOD. Funding Information: Acknowledgements. This material is based upon work supported, in part, by American Heart Association Grant 19IPLOI34760294 (to D.B.E.) and National Heart, Lung, and Blood Institute Grants R01-HL131823 (to D.B.E.), R01-HL152256 (to D.B.E.), and K25-HL135408 (to L.E.P.) and by the National Science Foundation under Grants 2205043 (to L.E.P.) and 2205103 (to D.B.E.). Funding Information: Acknowledgement. The authors gratefully acknowledge support from the French Agence Nationale de la Recherche (ANR) (grant ANR-22-CE45-0014-01, project MIRE4VTach), from the Atrial Fibrillation Chair of the IHU Liryc, from the Fon-dation Bordeaux Université, from the Fondation Lefoulon-Delalande, and from the French Federation of Cardiology - Grands projets - 2022 (project DIELECTRIC). Funding Information: Acknowledgments. This work was funded by British Heart Foundation Grants RE/13/4/30184 and RG/19/1/34160. Funding Information: Acknowledgements. We are grateful for the funding provided by the British Heart Foundation (ref: PG/22/10930), and the UK Engineering and Physical Sciences Research Council (EP/S030875, EP/S020950/1, EP/S014284/1, EP/R511705/1). Funding Information: Science Foundation under Grant Number 2205043. Funding Information: Ethical Considerations and Acknowledgements. This virtual study was carried out using computer simulations which did not require ethical approval. This research has been conducted using the UK Biobank Resource under Application Number 40161. The authors express no conflict of interest. This work was funded by an Engineering and Physical Sciences Research Council doctoral award, a Wellcome Trust Fellowship in Basic Biomedical Sciences (214290/Z/18/Z), the CompBioMed2 Centre of Excellence in Computational Biomedicine (European Commission Horizon 2020 research and innovation programme, grant agreement No. 823712). The computation costs were incurred through a PRACE ICEI project (icp019), which provided access to Piz Daint at the Swiss National Supercomputing Centre, Switzerland. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) public copyright license to any Author Accepted Manuscript (AAM) version arising from this submission. Funding Information: supported by NSF 2205103 and NIH R01 Funding Information: Acknowledgments. This work was supported by the European High-Performance Computing Joint Undertaking EuroHPC under grant agreement No 955495 (MICROCARD) co-funded by the Horizon 2020 programme of the European Union (EU) and the Swiss State Secretariat for Education, Research and Innovation. Funding Information: Acknowledgements. This work was supported by TUBITAK (grant no: 120N200), SAS (grant no: 536057), and VEGA (grant no: 2/0109/22). Funding Information: Acknowledgements. This work was supported by the Ministry of Education, Funding Information: Research reported in this publication was supported by the National Institute of Biomedical Imaging and Bioengineering of the National Institute of Health under award number R01EB027774. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The concepts and information Funding Information: Acknowledgements. This research was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research innovation programme under grant agreement no. 874827 (BRAV ). Funding Information: Acknowledgements. We gratefully acknowledge the financial support from the Health Research Council of New Zealand (17/608). We also acknowledge the important roles of our research nurses Mariska Oakester Bals, Jane Hannah, Anna Taylor, and Gracie Hoskin for their invaluable assistance in participant recruitment and data collection. Funding Information: Acknowledgements. The authors would like to thank Michel Ovize, Thomas Bocha-ton, and Nathan Mewton for sharing in-vivo data from the HIBISCUS cohort. We also thank Circle Cardiovascular Imaging (Calgary, Canada) for making the CVI42 software package available for research purposes. We acknowledge the support of the French Agence Nationale de la Recherche (ANR) under grants ANR-19-CE45-0020 (SIMR project), ANR-11-LABX-0063 (LABEX PRIMES of Univ. Lyon), and ANR-19-CE45-0005 (MIC-MAC project), and the Fédération Francaise de Cardiologie (MI-MIX project, Allocation René Foudon). Funding Information: Acknowledgements. This material is based upon work supported by the National Institutes of Health grants HL129077, HL119297 to MSS and RCG, and an American Heart Association pre-doctoral fellowship to NTS. Funding Information: Acknowledgements. This work is funded by EPSRC Centre for Doctoral Training in Smart Medical Imaging (EP/S022104/1), by a Program Grant from the British Heart Foundation (RG/19/1/34160), and Siemens Healthineers. This work was supported by the National Institute of Health (NIH R01-HL131823). Funding Information: Acknowledgements. This research was supported by the Innovate UK (104691) London Medical Imaging & Artificial Intelligence Centre for Value Based Healthcare, the USA National Institutes of Health R01HL121754, and core funding from the Wellcome/EPSRC Centre for Medical Engineering [WT203148/Z/16/Z]. SCOT-HEART was funded by The Chief Scientist Office of the Scottish Government Health and Social Care Directorates (CZH/4/588), with supplementary awards from Edinburgh and Lothian’s Health Foundation Trust and the Heart Diseases Research Fund. MCW was supported by the British Heart Foundation FS/ICRF/20/26002 and CH/09/002. SEW is supported by the British Heart Foundation (FS/20/26/34952). The authors acknowledge the support of the British Heart Foundation Centre for Research Excellence Award III (RE/18/5/34216). Funding Information: Acknowledgements. We thank and acknowledge VINNOVA (2022-00849), Digital Futures, and the Digital platform of KTH for their financial support; Dr. Marianne Schmid Daners and Dr. Thomas Gwosch at ETH Zurich for sharing their know-how in replicating the hybrid mock circulation loop and provision of the colacino-model implementation; Sara Mettler for the electrical support; Peter Arfert for the mechanical design; Laura Andersson, and Roxanne Rais for their experimental support. Funding Information: Acknowledgment. Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award No. R35GM128877 and by the Office of Advanced Cyberinfrastructure of the National Science Foundation under Award No. 1808530 and Award No. 1808553. Funding Information: Acknowledgments. This work was financially supported by the Theo Rossi di Mon-telera Foundation, the Metis Foundation Sergio Mantegazza, the Fidinam Foundation, and the Horten Foundation to the Center for CCMC. SP also acknowledges the CSCS-Swiss National Supercomputing Centre (No. s1074). Finally, this work was supported by the European High-Performance Computing Joint Undertaking EuroHPC under grant agreement No. 955495 (MICROCARD) co-funded by the Horizon 2020 programme of the European Union (EU) and the Swiss State Secretariat for Education, Research and Innovation. Funding Information: Acknowledgements. This work is funded in part by the 4TU Precision Medicine programme supported by High Tech for a Sustainable Future, a framework commissioned by the four Universities of Technology of the Netherlands. Jelmer M. Wolterink was supported by the NWO domain Applied and Engineering Sciences VENI grant (18192). This work made use of the Dutch national e-infrastructure with the support of the SURF Cooperative using grant no. EINF-2675. Funding Information: Acknowledgements. Chinese Heilongjiang Postdoctoral Grant (LBH-Z22184), National Natural Science Foundation of China (no. 52075133) and Metislab (Medical Engineering and Theory in Imaging and Signal Laboratory), CNRS LIA no.1124, INSA Lyon supported this work. This work was also performed within the framework of the LABEX PRIMES (ANR-11-LABX-0063) of Université de Lyon, within the program “Investissements d’Avenir” (ANR-11-IDEX-0007) operated by the French National Research Agency (ANR). Funding Information: The FIMH 2023 conference was certified by the MICCAI society and the French Society of Biomechanics (with financial support). Funding Information: Acknowledgements. This work was funded by Generalitat Valenciana Grant AICO/2021/318 (Consolidables 2021) and Grant PID2020-114291RB-I00 funded by MCIN/https://doi.org/10. 13039/501100011033 and by “ERDF A way of making Europe”. Funding Information: Acknowledgement. This work was supported in part by the British Heart Foundation, UK (Grant no. RG/F/22/110059). J Jevsikov is supported by the Vice Chancellor’s Scholarship at the University of West London. We are grateful to the following experts for their invaluable input in labelling images: Arjun Ghosh, Maysaa Zetani, Mahmoud Tawil, Luxy Ananthan, Camelia Demetrescu, Amar Singh, Sanjeev Bhattacharyya, Joban Sehmi, Kavitha Vimalesvaran, Abdallah Al-Mohammad, Bushra Rana, Tiffany Ng. Funding Information: Acknowledgements. This work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1-DTH-06-2020; grant agreement number 101016496). This study received financial support from the French Government as part of the “Investments of the Future” program managed by the National Research Agency (ANR), Grant reference ANR-10-IAHU-04. Experiments presented in this paper were partially carried out using the PlaFRIM experimental testbed, supported by Inria, CNRS (LABRI and IMB), Université de Bordeaux, Bordeaux INP and Conseil Régional d’Aquitaine. Funding Information: Acknowledgement. This research has been conducted using the UK Biobank Resource under Application Number ‘40161’. The authors express no conflict of interest. This work was funded by the CompBioMed 2 Centre of Excellence in Computational Biomedicine (European Commission Horizon 2020 research and innovation programme, grant agreement No. 823712). L. Li was partially supported by the SJTU 2021 Outstanding Doctoral Graduate Development Scholarship. A. Banerjee is a Royal Society University Research Fellow and is supported by the Royal Society Grant No. URF\R1\221314. The work of A. Banerjee and V. Grau was partially supported by the British Heart Foundation (BHF) Project under Grant PG/20/21/35082. Funding Information: Acknowledgements and Funding. All staff from LIRYC and CHU Bordeaux involved in the Human donor program CADENCE and HARMONICA project are gratefully acknowledged for their valuable contributions. This work received financial support from the French National Investments for the Future Programs: ANR-10-IAHU-04. HD figures are available at: https://github.com/ valeryozenne/Cardiac-Structure-Database/tree/master/Article-4. Publisher Copyright: © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

PY - 2023/6/16

Y1 - 2023/6/16

N2 - Displacement ENcoding with Stimulated Echoes (DENSE) is a CMR modality that can encode myocardial tissue displacement at a pixel level, enabling the characterization of cardiac disease at early stages. However, we do not currently have a way of evaluating the accuracy of derived results, since the ground truth is unknown. In this study, we developed a proof-of-concept pipeline to generate realistic DENSE images with a known ground truth. We leverage the XCAT tool to create body anatomies, along with associated myocardial tissue displacements, and generate DENSE images with a Bloch simulation based on the time-resolved positions. We generated 6 samples: an apical, a mid, and a basal short-axis slice for both male and female anatomy. We then extracted radial and circumferential strain components using DENSEanalysis, and compared them to the ground-truth strain obtained from the XCAT displacements. While the reproducibility of the strain calculations was similar to the inter-observer variability from previous studies, and the bias in circumferential strain was small (0.03 ± 0.02), the current methods for strain extraction resulted in a bias in radial strain of 0.19 ± 0.19. There is a need to develop better regularization strategies for DENSE analysis, for instance using Deep Learning, and this study provides initial groundwork for obtaining ground-truth strain to evaluate these methods.

AB - Displacement ENcoding with Stimulated Echoes (DENSE) is a CMR modality that can encode myocardial tissue displacement at a pixel level, enabling the characterization of cardiac disease at early stages. However, we do not currently have a way of evaluating the accuracy of derived results, since the ground truth is unknown. In this study, we developed a proof-of-concept pipeline to generate realistic DENSE images with a known ground truth. We leverage the XCAT tool to create body anatomies, along with associated myocardial tissue displacements, and generate DENSE images with a Bloch simulation based on the time-resolved positions. We generated 6 samples: an apical, a mid, and a basal short-axis slice for both male and female anatomy. We then extracted radial and circumferential strain components using DENSEanalysis, and compared them to the ground-truth strain obtained from the XCAT displacements. While the reproducibility of the strain calculations was similar to the inter-observer variability from previous studies, and the bias in circumferential strain was small (0.03 ± 0.02), the current methods for strain extraction resulted in a bias in radial strain of 0.19 ± 0.19. There is a need to develop better regularization strategies for DENSE analysis, for instance using Deep Learning, and this study provides initial groundwork for obtaining ground-truth strain to evaluate these methods.

KW - Cardiac MR

KW - DENSE

KW - Simulation

KW - Strain

UR - http://www.scopus.com/inward/record.url?scp=85172737058&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-35302-4_43

DO - 10.1007/978-3-031-35302-4_43

M3 - Chapter

AN - SCOPUS:85172737058

SN - 9783031353017

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 412

EP - 421

BT - Functional Imaging and Modeling of the Heart

A2 - Bernard, Olivier

A2 - Clarysse, Patrick

A2 - Duchateau, Nicolas

A2 - Ohayon, Jacques

A2 - Viallon, Magalie

PB - Springer Nature

CY - Cham

T2 - Functional Imaging and Modeling of the Heart - 12th International Conference, FIMH 2023, Proceedings

Y2 - 19 June 2023 through 22 June 2023

ER -