Markus Henningsson^1
1Linköping University, Linköping, Sweden

A method for simultaneous T₁ and T₂ mapping of the myocardium is proposed, termed Multimapping, based on dictionary matching using a simple Cartesian single-shot acquisition across 10 cardiac cycles. The method is evaluated in a phantom, 12 healthy subjects and 43 patients with suspected cardiomyopathy. 

Maxim Terekhov¹, David Lohr¹, Michael Hock¹, Maya Bille¹, Julia Aures¹, Ibrahim A. Elabyad¹, Florian Schnitter², Wolfgang Bauer², Ulrich Hofmann², and Laura M. Schreiber^1
1Chair of Cellular and Molecular Imaging, Comprehensive Heart Failure Center, University Hospital Würzburg, Comprehensive Heart Failure Center, Wuerzburg, Germany, ²Department of Internal Medicine I, Cardiology, University Hospital Würzburg, Wuerzburg, Germany

Cardiac MRI with T₂^*-contrast at 7T  is of particular interest because it is highly sensitive to cardiac tissue alterations after myocardial injury. A reliable T₂^*-contrast quantification and usage for tissue characterization require minimization of the inhomogeneity of the B₀-field in the heart. In this work, we analyzed the  B₀-conditions achievable in longitudinal  T₂^*  measurements in the hearts of domestic pigs which during the experiment grew from 30 to 90 kg.  B₀ distribution statistics were summarized for the maps of individual slices.  Demands on the 7T-scanners B₀-shimming hardware (up to 3rd-order spherical harmonics) were analyzed.  

Seung Su Yoon^1,2, Michaela Schmidt², Manuela Rick², Teodora Chitiboi³, Puneet Sharma³, Tilman Emrich^4,5, Christoph Tillmanns⁶, Andreas Seitz⁷, Heiko Mahrholdt⁷, Solenn Toupin⁸, Théo Pezel^9,10, Jérôme Garot⁹, Jens Wetzl², and Andreas Maier^1
1Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, ²Magnetic Resonance, Siemens Healthcare GmbH, Erlangen, Germany, ³Siemens Medical Solutions USA, Inc, Princeton, NJ, United States, ⁴Department of Radiology, University Medical Center, Johannes Gutenberg-University Mainz, Mainz, Germany, ⁵Division of Cardiovascular Imaging, Department of Radiology and Radiological Science, Medical University of South Carolina, Charleston, Germany, ⁶Diagnostikum Berlin, Berlin, Germany, ⁷Department of Cardiology, Robert Bosch Medical Center, Stuttgart, Germany, ⁸Siemens Healthcare France, Saint-Denis, France, ⁹Institut Cardiovasculaire Paris Sud, Cardiovascular Magnetic Resonance Laboratory, Hôpital Privé Jacques Cartier - Ramsay Santé, Massy, France, ¹⁰Division of Cardiology, Johns Hopkins University, Baltimore, MD, United States

In cardiac MRI, the Late Gadolinium Enhancement technique is usually performed after inversion recovery scout sequences that are acquired to null the myocardial properly for optimal image contrast. In clinical practice, the selection and adjustment of the inversion time for healthy myocardium nulling are manually performed. To standardize and automate the process, we propose an automated deep-learning-based system combined with a linear regression model, which automates the selection of the inversion time, taking into consideration the time delay between the inversion time scout and LGE sequences. In this work, we validated the system in a large retrospective study (N=765).

Carmel Hayes¹, Yan Tu Huang², Manuela Rick¹, Randall Kroeker¹, Mario Bacher¹, Yue Pan³, Juliet Varghese³, Orlando Simonetti³, Ning Jin⁴, Rachel Davids⁴, Kelvin Chow⁴, Rudy Vanliedekerke ⁵, Marieke Vangelder⁵, Johan Dehem⁵, Peter Gatehouse⁶, Raj Kumar Soundarajan ⁶, Ricardo Wage⁶, Alessia Azzu⁶, Sonia Nielles-Vallespin⁶, and Peter Speier^1
1Siemens Healthcare GmbH, Erlangen, Germany, ²Siemens Shenzhen Magnetic Resonance, Shenzhen, China, ³The Ohio State University, Columbus, OH, United States, ⁴Siemens Medical Solutions USA, Malvern, PA, United States, ⁵Jan Yperman Ziekenhuis, Ieper, Belgium, ⁶Royal Brompton Hospital, London, United Kingdom

A multi-center evaluation of a Pilot Tone-based triggering technique using the BioMatrix Beat Sensor has been performed in healthy volunteers and patients undergoing a cardiac MRI examination. At five different sites, and two different field strengths, a wide range of MR sequences typically used in standard cardiac MRI examinations were successfully triggered using the prototype Beat Sensor Cardiac triggering technology.  By including a short signal training and RF calibration step at the start of the examination, the device proved robust across a range of subjects, thereby offering an alternative means to trigger cardiac examinations without the need to attach electrodes.

Junwei Yang^1,2, Pengfang Qian², Thomas Kuestner³, Pietro Lio¹, Dinggang Shen^2,4, and Haikun Qi^2
1Department of Computer Science and Technology, University of Cambridge, Cambridge, United Kingdom, ²School of Biomedical Engineering, ShanghaiTech University, Shanghai, China, ³Department of Interventional and Diagnostic Radiology, University Hospital of Tuebingen, Tuebingen, Germany, ⁴Shanghai United Imaging Intelligence Co., Ltd., Shanghai, China

Numerous deep learning methods have been proposed for cardiac cine MRI segmentation, while most of them require laborious annotation for supervised training.  Herein, we propose an approach to perform group-wise registration and joint segmentation of cardiac cine images, by training a registration network in a self-supervised manner to align dynamic images to their mean image space and also a segmentation network in a weakly-supervised manner using sparsely-annotated data and predicted motions from the registration network. By training these two (registration and segmentation) networks simultaneously, our proposed joint learning approach provides better segmentations than the direct segmentation network.

Pamela Franco^1,2,3, Julio Sotelo^1,3,4, Andrea Guala⁵, Lydia Dux-Santoy⁵, Arturo Evangelista⁵, José Rodríguez-Palomares⁵, Domingo Mery⁶, Rodrigo Salas⁴, and Sergio Uribe^1,3,7
1Biomedical Imaging Center, School of Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, ²Electrical Engineering Department, School of Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, ³Millennium Nucleus for Cardiovascular Magnetic Resonance, Santiago, Chile, ⁴School of Biomedical Engineering, Universidad de Valparaíso, Valparaíso, Chile, ⁵Department of Cardiology, Hospital Universitari Vall d’Hebron, Vall d’Hebron Institut de Recerca (VHIR), Universitat Autònoma de Barcelona, Barcelona, Spain, ⁶Deparment of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile, ⁷Radiology Department, School of Medicine, Pontificia Universidad Católica de Chile, Santiago, Chile

Several studies have demonstrated the existence of altered hemodynamics in bicuspid aortic valve (BAV) patients. The objective of this study was to identify which hemodynamic parameters allow an accurate classification between BAV patients with dilated and non-dilated ascending aorta using machine learning (ML) algorithms.

Marco Barbero Mota¹, Giacomo Annio¹, Guillaume Rucher², Anna Wittgestein³, David Nordsletten^3,4, Jordi Martorell⁵, and Ralph Sinkus^1,3
1LVTS U1148, LVTS U1148 INSERM - Université de Paris, Paris, France, ²FRIM, INSERM - Université de Paris, Paris, France, ³School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, ⁴Biomedical Engineering and Cardiac Surgery, University of Michigan, Ann Arbour, MI, United States, ⁵Chemical Engineering, IQS, Barcelona, Spain

Heart biomechanics play a crucial role in the diagnosis and prognosis of cardiac diseases. However, current gold standard methods are highly invasive and preclude continuous monitoring of patients’ myocardium stiffness as a biomarker. In this study, we report the first results of a novel magnetic resonance elastography technique that provides with high temporal and spatial resolution enabling accurate transient shear wave speed values quantification, as thereof travel through heart tissue. Although biased due to the thin-plate geometry of the heart, early and mid-late systole, and early diastole preliminary results intuitively match the expected dynamical biomechanics of the human heart.

Jack J. Miller^1,2,3, Matthew K Burrage², Moritz Hundertmark², Ladislav Valkovič^2,4, William D Watson², Jennifer Rayner², Nikant Sabharwal⁵, Vanessa M Ferreira², Stefan Neubauer^2,5, Oliver J Rider^2,5, and Andrew J Lewis^2,5
1Aarhus University, Aarhus, Denmark, ²OCMR (Oxford Center for Clinical Magnetic Resonance Research), University of Oxford, Oxford, United Kingdom, ³Department of Physics, University of Oxford, Oxford, United Kingdom, ⁴Department of Imaging Methods, Slovak Academy of Sciences, Bratislava, Slovakia, ⁵Department of Cardiology, Oxford University Hospitals NHS Trust, Oxford, United Kingdom

We show in 43 patients with ³¹P MRS, 1H CMR and a novel proton-density mapping sequence that a gradient of myocardial energetic impairment exists across a spectrum of HFpEF phenotypes of increasing clinical severity and worsening diastolic function. A greater degree of myocardial energetic deficit is linked to impaired LV systolic and diastolic functional reserve, to altered RV reserve and RV-PA coupling, and to exercise-induced pulmonary congestion assessed using novel proton density magnetic resonance imaging. A subgroup of HFpEF patients demonstrate transient pulmonary congestion during exercise which can be non-invasively assessed using exercise CMR and proton density mapping.

Marco Marino^1,2, Nigel Colenbier^1,2,3, Nicola Filippini², Giovanni Pellegrino², Daniele Marinazzo^2,3, and Giulio Ferrazzi^2
1Research Center for Motor Control and Neuroplasticity, KU Leuven, Leuven, Belgium, ²IRCCS San Camillo Hospital, Venice, Italy, ³Department of Data Analysis, Faculty of Psychology and Educational Sciences, Ghent University, Ghent, Belgium

Cardiac pulsation is a physiological confound of fMRI analysis pipelines and so accurate mapping of cardiac contributions to the BOLD signal is highly warranted. To overcome cardiac aliasing associated with the limited temporal resolution in fMRI, we developed a data-driven methodology to spatially and temporally resolve cardiac contributions from the BOLD signal itself (i.e without the need of processing external physiological recordings such as PPU/ECG signals). This is achieved by combining simultaneous multi-slice imaging and a dedicated hyper-sampling decomposition scheme.

The proposed methodology is fully data driven and it does not make specific assumptions on the shape of cardiac pulsation.

Žiga Bizjak¹, Aichi Chien, PhD, FAHA², Erazem Kos¹, and Žiga Špiclin^1
1Laboratory of Imaging Technologies, Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia, ²Division of Interventional Neuroradiology, Department of Radiological Sciences, Ronald Reagan UCLA Medical Center, David Geffen School of Medicine at UCLA, Los Angeles, CA, United States

Management of (un)ruptured intracranial aneurysms (IAs) through early detection and continuous rupture risk assessment is becoming an integral part of IA treatment. This study showcases the capabilities of AVA Diagnostics tool and its IA workflow including automatic vessel segmentation, detection and isolation, morphologic quantification, future growth prediction and follow-up morphologic analysis. Each of the tools was successfully validated on multimodal MRA, CTA and 3D-DSA scans. The AVA Diagnostics thus aims to simplify the management of IAs through extensive use of computer-assisted image analysis and analytics tools, and enable their application in large-scale studies and in regular clinical workflow.

Shi-Ming Wang¹, Guglielmo Genovese^2,3,4, Belen Diaz^3,5, Fan Huang⁶, Stéphane Lehericy^2,3, Hui Zhang¹, Charlotte Rosso^3,5, Francesca Branzoli^2,3, and Marco Palombo^1,7,8
1Centre for Medical Image Computing (CMIC), Department of Computer Science, University College London, London, United Kingdom, ²Paris Brain Institute - ICM, Centre de NeuroImagerie de Recherche - CENIR, Paris, France, ³Sorbonne Université, UMR S 1127, Inserm U 1127, CNRS UMR 7225, ICM, F-75013, Paris, France, ⁴Center for Magnetic Resonance Research and Department of Radiology, University of Minnesota, Minneapolis, MN, United States, ⁵Department of Neurology, Pitié-Salpétrière Hospital, Paris, France, ⁶Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, ⁷Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, United Kingdom, ⁸School of Computer Science and Informatics, Cardiff University, Cardiff, United Kingdom

This study characterises microstructural changes in stroke lesions using NODDI and WMTI, two advanced diffusion MRI techniques, and compares their ability to detect WM alterations on the ipsi-lesional and contralesional sides of 12 patients scanned 2 weeks, 1 month and 3 months after onset. Region-of-interest analysis showed that ODI and FWF from NODDI were significantly altered at the lesion. Significant changes were also found in the WMTI parameters of several ipsi-lesional white matter tracts. Our prospective analysis suggests that the application of enhanced dMRI methods such as NODDI and WMTI can help to further comprehend the biophysical mechanism behind ischemia.

Esra Zihni¹, Bryony L. McGarry^1,2, Jen Guo³, Rani Gupta Sah³, George Tadros³, Philip A. Barber³, and John D. Kelleher^1,4
1PRECISE4Q Predictive Modelling in Stroke, Technological University Dublin, Dublin, Ireland, ²School of Psychological Science, University of Bristol, Bristol, United Kingdom, ³Calgary Stroke Program, Department of Clinical Neurosciences, University of Calgary, Calgary, AB, Canada, ⁴ADAPT Research Centre, ICE Research Institute, Technological University Dublin, Dublin, Ireland

Applying deep learning models to MRI scans of acute stroke patients to extract features indicative of functional outcome could assist a clinician’s treatment decisions. Here, we trained convolutional neural network models on ADC maps from hyper-acute ischemic stroke patients to predict 3-month mRS and used an interpretability technique to highlight regions in the ADC maps that were most important in the prediction of good and poor outcomes. Although the models had poor predictive power, the visual explanations supported our previous findings that predictions might be based not on ischemic regions, but on other relevant information inherent in the image.

Žiga Bizjak¹, Aichi Chien, PhD, FAHA², Jan Tasič¹, and Žiga Špiclin^1
1Laboratory of Imaging Technologies, Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia, ²Division of Interventional Neuroradiology, Department of Radiological Sciences, Ronald Reagan UCLA Medical Center, David Geffen School of Medicine at UCLA, Los Angeles, CA, United States

The correlation between different variants of  the Circle of Willis (CoW) and cerebrovascular disorders such as stroke, aneurysms and mental disorders is not yet well understood. A step towards a better understanding of the role of CoW in the aforementioned diseases is the automatic labeling of main vessels. In this work we tested three different approaches and observed high mIoU value of 0.870. As such the automatic anatomical labelling of the CoW seems feasible for clinical evaluation of the association of different anatomical variants with the risk factors of cerebrovascular pathologies.