Simon Reiss¹, Johannes Fischer¹, Julien Thielmann², Thomas Lottner¹, Timo Heidt², Constantin von zur Mühlen², and Michael Bock^1
1Dept. of Radiology, Medical Physics, Medical Center University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany, ²Dept. of Cardiology and Angiology I, University Hospital Freiburg and Faculty of Medicine, University of Freiburg, Freiburg, Germany

Magnetic resonance provides a multitude of imaging techniques to detect and characterize myocardial infarction (MI). Terminal animal studies are often performed in which CMR findings after MI are correlated to histology. However, precise manual extraction of tissue samples from a specific myocardial region defined by CMR is difficult. Here, we propose the use of 3D-printed guides for extraction of tissue samples from myocardial regions pre-defined on high resolution ex vivo CMR. Using the presented technique, histology findings can be correlated to the exact position within the MRI data set with a precision of < 1 mm.

Arjun Gupta¹, Labonny Biswas², Jay Soni², Brandon Coles¹, Sebastian Ferguson², Graham A. Wright³, and Nilesh R. Ghugre^3
1Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, ²Physical Science Platform, Sunnybrook Research Institute, Toronto, ON, Canada, ³Schulich Heart Program, Sunnybrook Research Institute, Toronto, ON, Canada

Minimally invasive cardiac interventions, such as ablation treatment for ventricular tachycardia and therapy delivery involving biologics (cells, genes, biomaterials) for the treatment of myocardial infarction, require a tracking accuracy of <5mm. Our study aimed to characterize the spatial error associated with MR-guided active catheter tracking under static conditions. In a 1.5T MR scanner, we tracked an ablation catheter at 48 different positions, all confined within the boundaries of where a typical human heart would be located. We found that the error was both minimized and maintained below our 5mm target when using a Hadamard multiplexed active tracking sequence.

Gareth R Davies¹, Robert S Stormont², Lionel M Broche¹, Dana K Dawson¹, David J Lurie¹, and P James Ross^1
1Aberdeen Biomedical Imaging Centre, University of Aberdeen, Aberdeen, United Kingdom, ²GE Healthcare, Waukesha, WI, United States

The SNR and inter coil coupling properties of a range of litz wire coils, in conjunction with the circuitry to match the coils to a low impedance pre-amplifier, were examined. The results of these studies were combined with observations of SNR and coil coupling found in the images of phantoms, obtained with various array designs, to construct an RF array optimized for 0.2 T (8.5 MHz). The array has demonstrated its capability in acquiring cardiac images from human volunteers in our field cycling imager.

Belinda Ding¹, Liam A.J. Young^1,2, Margaret Huang^3,4, Aravinda Page⁵, Simon Messer⁵, Mike Murphy³, Justin Perkins⁶, Elizabeth Tunnicliffe², Stephen Large⁵, Christopher T Rodgers¹, and Jonathan R Weir-McCall^5,7
1Wolfson Brain Imaging Centre, University of Cambridge, Cambridge, United Kingdom, ²Oxford Centre for Magnetic Resonance, University of Oxford, Oxford, United Kingdom, ³MRC Mitochondrial Unit, University of Cambridge, Cambridge, United Kingdom, ⁴Department of Surgery, University of Cambridge, Cambridge, United Kingdom, ⁵Royal Papworth Hospital, Cambridge, United Kingdom, ⁶Royal Veterinary College, London, United Kingdom, ⁷Department of Radiology, University of Cambridge, Cambridge, United Kingdom

Ischaemia-reperfusion injury is a key pathophysiological process driving cardiac injury in myocardial infarction, heart surgery and transplants. Succinate accumulation is a specific marker of ischaemic injury, and a potential therapeutic target to reduce reperfusion injury.Single-voxel proton spectra were acquired during normal perfusion and during ischaemia in three machine-perfused porcine hearts. During ischaemia, a peak was evident in the expected region of succinate, which was absent during normal perfusion. This was further examined in a phantom model which confirmed the peak location with concentration-signal correlation of R>0.999.  MRS detection of succinate holds potential for the detection of ischaemia.

Margaret Samyn¹, Ke Yan², Kristen George-Durrett³, Justin Godown³, Kimberly Crum³, Maryanne Christ⁴, Kak-Chen Chan⁴, Kak-Chen Chan⁴, David Bearl³, Debra Dodd³, Lazaro Hernandez⁴, Jonathan Soslow³, and Bruce Damon^5
1Pediatrics, Medical College of Wisconsin, Milwaukee, WI, United States, ²Biostatistics, Medical College of Wisconsin, Milwaukee, WI, United States, ³Pediatrics, Division of Pediatric Cardiology, Vanderbilt University Medical Center, Nashville, TN, United States, ⁴Pediatric Cardiology, Joe DiMaggio Children’s Hospital at Memorial Healthcare System, Hollywood, FL, United States, ⁵Stephens Family Clinical Research Institute Carle Foundation Hospital, Urbana, IL, United States

Acute rejection (AR) continues to cause significant morbidity and mortality in pediatric heart transplant recipients (PHTx). Endomyocardial biopsy is the standard-of-care for diagnosis of AR, but it is invasive and associated with morbidity and mortality and can miss patchy AR. Cardiac magnetic resonance (CMR) was performed using parametric mapping and novel texture analysis to detect patterns of myocardial edema and fibrosis.  Patients with AR had significant differences in texture analysis compared with patients without AR.  CMR with texture analysis has potential as a non-invasive method for detection of AR in PHTx.

Akos Varga-Szemes¹, Teodora Chitiboi², U. Joseph Schoepf¹, Athira J Jacob², Sayan Ghosal³, Fei Xiong⁴, Puneet Sharma², Jonathan Aldinger¹, and Tilman Emrich^1
1Medical University of South Carolina, Charleston, SC, United States, ²Siemens Healthcare, Princeton, NJ, United States, ³Johns Hopkins University, Baltimore, MD, United States, ⁴Siemens Healthcare, Charleston, SC, United States

Existing work demonstrates the value of image-based cardiovascular diagnosis with AI. We aimed to develop and test a machine learning algorithm for cardiovascular disease classification based on cine image datasets of 570 consecutive patients. Disease classification was performed using a random forest (RF) classifier and a disease classification network based on graph attention networks. The fully automated deep learning algorithm showed high accuracy for cardiac disease classification based on cine images only. Such algorithm has the potential to improve the efficiency of the reading process, especially by identifying and filtering out patients with normal cardiac anatomy and function.

Ana Fehrmann¹, David Maintz¹, and Bettina Baeßler^2
1Department of Radiology, University Hospital of Cologne, Cologne, Germany, ²Department of Radiology, University Hospital of Würzburg, Würzburg, Germany

Due to demographic changes, the prevalence of cardiac conditions predisposing to diastolic dysfunction (DD) is rising steadily. The main focus of non-invasive assessment of parameters for DD has been on left ventricular strain alterations. In this prospective study, we performed strain analysis of all 4 cardiac chambers using cardiac magnetic resonance feature tracking in patients with echocardiographically diagnosed DD and compared the results to a healthy control group. Left atrial strain parameters proved to be the most potent predictors of DD and allowed for a clear separation between patients and age-matched controls.

Céline Marquet^1,2,3, Jihye Jang^1,2,4, Andrew J. Powell^1,2, Brian Fonseca^5,6, Lorna P. Browne^5,6, and Mehdi H. Moghari^5,6
1Harvard Medical School, Boston, MA, United States, ²Boston Children’s Hospital, Boston, MA, United States, ³Technical University of Munich, Munich, Germany, ⁴Philips Healthcare, Boston, MA, United States, ⁵University of Colorado, Aurora, CO, United States, ⁶Children’s Hospital Colorado, Aurora, CO, United States

We developed a novel method (RavFa-T₂) for T₂ quantification of myocardium using transient steady-state free precession (SSFP) with a variable flip angle scheme. RavFa-T₂ was systematically analyzed based on a phantom and a two-center patient study. We show that RavFa-T₂ yields accurate T₂ estimates for the myocardium compared to the state-of-the-art method T₂-Prep and reduces scan time to ≤4 seconds.

Ignasi Alemany^1,2, Jan N Rose¹, Pedro F. Ferreira^2,3, Sonia Nielles-Vallespin^2,3, Denis J. Doorly¹, and Andrew D. Scott^2,3
1Department of Aeronautics, Imperial College of London, London, United Kingdom, ²Cardiovascular Magnetic Resonance Unit, Royal Brompton Hospital, London, United Kingdom, ³National Heart and Lung Institute, Imperial College London, London, United Kingdom

Diffusion tensor cardiovascular magnetic resonance (DT-CMR) has proved to be a unique non-invasive method to investigate the myocardial microstructure. The time dependence of the diffusion tensor provides insights into the barriers that restrict free diffusion. We present a method to assess the time dependence of diffusion in the beating heart along with numerical simulations that demonstrate the sensitivity of this potential new biomarker to changes in myocardial microstructure including membrane permeability, extracellular volume fraction and cell hypertrophy. 

Jonathan Weine¹, Stefano Buoso¹, and Sebastian Kozerke^1
1Institute for Biomedical Engineering, University and ETH Zurich, Zürich, Switzerland

Cardiac diffusion tensor imaging (cDTI) as well as cardiac strain imaging both offer invaluable information on the functional state of the heart. Acquiring both, takes too long to be clinically feasible. Free breathing, cDTI acquisition with second-order motion compensated diffusion encoding waveforms display residual phase variations, that encode higher orders of contractile motion during the diffusion encoding process. This work presents a theoretical framework on how to jointly infer strain and diffusion tensors from complex-valued cDTI data with varying trigger delays.

Irvin Teh¹, Christopher Kelly¹, David Shelley², Ana-Maria Poenar¹, Sven Plein¹, Erica Dall'Armellina¹, Christopher Nguyen^3,4, and Jürgen E. Schneider^1
1Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, United Kingdom, ²Leeds Teaching Hospitals Trust, Leeds, United Kingdom, ³Massachusetts General Hospital, Harvard Medical School, Cardiovascular Research Center, Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ⁴Health Science Technology, Harvard-MIT, Cambridge, MA, United States

There is strong interest in characterising the cardiac microstructure using in vivo cardiac diffusion tensor imaging (CDTI). However, its use in larger clinical studies is often hampered by long scan times. We sought to rationalise the scan parameters needed for a clinically feasible CDTI protocol, by comparing carefully subsampled data against a 24-minute reference dataset. A design strategy was identified based on maximising the number of diffusion-weighting (DW) directions, subject to minimum SNR requirements. Feasibility of a 5-minute protocol was demonstrated where NRMSE(MD) = 5.2±0.2%, NRMSE(FA) = 12.8±0.2%, RMSE(HA) = 5.5±0.4°, RMSE(absE2A) = 15.7±1.9°.

Julie Magat^1,2,3, Maxime Yon^1,2,3, Yann Bihan-Poudec⁴, and Valéry Ozenne^1,2,3,5
1IHU Liryc, Electrophysiology and Heart Modeling Institute, Foundation Bordeaux Université, Bordeaux, France, ²Univ. Bordeaux, Centre de recherche Cardio-Thoracique de Bordeaux, U1045, Bordeaux, France, ³INSERM, Centre de recherche Cardio-Thoracique de Bordeaux, U1045, Bordeaux, France, ⁴Institut des Sciences Cognitives Marc Jeannerod, CNRS UMR 5229, Université Claude Bernard Lyon I, Bron, France, ⁵Centre de Résonance Magnétique des Systèmes Biologiques, UMR 5536, CNRS/Université de Bordeaux, Bordeaux, France

In this study, we developed a groupwise registration and tractography framework to investigate the global myofiber arrangement of large mammalian sheep hearts. To demonstrate the potential application of the proposed method, a novel description of sub-regions in the intraventricular septum (IVS) is presented. The study focuses on one fiber-bundle in the posterior junction and three fiber-bundles in the anterior junction.

Ben K Statton¹, Alaine Berry¹, Peter Kellman², Hui Xue², Stuart Cook³, and Declan O'Regan^3
1MRC, London Institute of Medical Sciences, Imperial College London, London, United Kingdom, ²National Heart, Lung, and Blood Institute, National Institute of Health, Bethesda, MD, United States, ³MRC, London Institute of Medical Sciences, UKRI, London, United Kingdom

Cardiac magnetic resonance with a retrospective ECG-gated breath-hold cine stack is the reference standard for assessing LV volumes, however this technique is unsuitable for patients who cannot hold their breath or suffer arrhythmia. A potential solution is a “retro-gated” real-time sequence which is prospectively triggered to acquire 120 frames over multiple heartbeats in each slice. Images from a beat most closely matching the median beat length for the entire stack are then temporally interpolated to 30 output phases for each slice.  This retro-gated method showed good agreement with the reference standard sequence in a cohort of healthy volunteers.

Bård Andre Bendiksen^1,2,3, Einar Sjaastad Nordén^1,2,3, Ivar Sjaastad^1,2, Lili Zhang^1,2, and Emil Knut Stenersen Espe^1,2
1Institute for Experimental Medical Research, University of Oslo and Oslo University Hospital, Oslo, Norway, ²KG Jebsen Center for Cardiac Research, University of Oslo, Oslo, Norway, ³Oslo New University College, Oslo, Norway

Little is known about the relationship between cardiac function and myocardial fiber structure after myocardial infarction (MI). We hypothesized that post-MI remodeling of left ventricular fiber structure would be reflected in changes in the regional strains. Strains and cardiac fiber structure were measured using MRI in five rats with MI, and three sham-operated rats. Transmural variation in fiber helix angles was reduced adjacent to the infarct. This reduction was associated with reduced radial, longitudinal, and endocardial circumferential strain, and improved epicardial circumferential strain in the adjacent zone. Our findings emphasize the relevance of cardiac fiber structure remodeling after myocardial infarction.