Qing Zou¹, Abdul Haseeb Ahmed¹, Prashant Nagpal², Sarv Priya¹, Rolf F. Schulte³, and Mathews Jacob^1
1University of Iowa, Iowa City, IA, United States, ²University of Wisconsin–Madison, Madison, WI, United States, ³GE Global Research, Munich, Germany

Free-breathing cardiac cine methods are needed for pediatric and chronic obstructive pulmonary disease (COPD) subjects. Multi-slice acquisitions can offer good blood-myocardium contrast, and hence preferred over 3D methods. Current approaches independently recover the slices, followed by post-processing to combine data from different slices. In this work, a deep manifold learning scheme is introduced for the joint alignment and reconstruction of multi-slice dynamic MRI. The proposed scheme jointly learns the parameters of the deep network as well as the latent vectors for each slice, which captures the motion induced dynamic variations, from the k-t space data of the specific subject. 

Thomas Lindner¹, Friederike Austein¹, Helena Guerreiro¹, and Jens fiehler^1
1Department of Diagnostic and Interventional Neuroradiology, University Hospital Hamburg-Eppendorf, Hamburg, Germany

This work presents a method of rapidly acquiring a T1 map to optimize background suppression in Arterial Spin Labeling to improve the signal-to-background contrast.

Michael G Crabb¹, Karl P Kunze^1,2, Carlos Velasco¹, Anastasia Fotaki¹, Camila Munoz¹, Alina Hua¹, Radhouene Neji^1,2, Claudia Prieto¹, and Rene M Botnar^1
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, ²MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom

In patients with suspected myocardial infarction, late-gadolinium enhancement (LGE) images are often acquired to detect focal and diffuse scarring. T1ρ is a sensitive marker for assessment of macromolecular-water interaction and has shown potential for non-contrast enhanced imaging of fibrosis. Here we propose a novel free-breathing, 3D whole-heart joint T1/T1ρ mapping sequence with Dixon encoding to provide 3D T1 and T1ρ maps and co-registered water and fat volumes with isotropic resolution for comprehensive contrast-agent free myocardial tissue characterization. Preliminary 3D joint T1/T1ρ mapping and water/fat imaging results demonstrate good agreement in phantoms and promising results in-vivo in healthy volunteers and patients.

Andrew Phair¹, Gastao Cruz¹, Haikun Qi², René Botnar¹, and Claudia Prieto^1
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, ²School of Biomedical Engineering, Shanghai Tech University, Shanghai, China

Recent work has enabled the simultaneous acquisition of 3D myocardial T₁ and T₂ maps with isotropic spatial resolution and cardiac cine images from a ~10-minute scan. Herein, we propose to incorporate non-rigid cardiac motion correction into a dictionary-based low-rank reconstruction scheme, allowing k-space data from all cardiac phases to be included in the reconstruction of any given phase. Reconstructed cine images and T₁ and T₂ maps with and without motion correction are presented and demonstrate that a reduction to 30% of the acquired k-space data (~3-minute scan) can be achieved while image quality is maintained.

Jie Zheng^1,2, Ran Li^1,2, Cihat Eldeniz¹, Thomas H Schindler¹, Linda R Peterson³, and Pamela K Woodard^1,2
1Mallinckrodt Institute of Radiology, Washington University in St. Louis, Saint Louis, MO, United States, ²Department of Biomedical Engineering, Washington University in St. Louis, Saint Louis, MO, United States, ³Department of Medicine, Cardiovascular Division, Washington University in St. Louis, Saint Louis, MO, United States

A previously developed MRI method for quantitative myocardial oxygen extraction mapping showed promising results, but image quality suffered from distortion and low in-plane resolution. Therefore we developed a new image acquisition method which both doubled the in-plane spatial resolution and corrected image distortion and tested it in healthy subjects. Reproducibility studies showed comparable results between the two methods. Rigorous animal and/or human validation studies are warranted to study its translational potential in the assessment of patients with myocardial metabolic dysfunction.

Chuan-Miao Xie¹, Jie Liu², Jia-Lei Zhao², Qi Liu³, Jian Xu³, Li-Yun Zheng⁴, Xin Liu², Hairong Zheng², and Yin Wu^2
1Sun Yat-Sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China, ²Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, ³UIH America, Inc., Houston, TX, United States, ⁴Central Research Institute, United Imaging Healthcare, Shanghai, China

This study investigated the dynamic alteration of myocardium creatine in infarct hearts using CEST MRI. Seven pigs underwent cine, Cr CEST and LGE imaging 3 and 14 days after MI. Significant increase of stroke volume and ejection fraction and decrease of infarct angle were observed at day 14 compared to that at day 3. Meanwhile, Cr CEST signals increased significantly in the infarct and its adjacent myocardium, and negatively correlated with infarct angle. The improved metabolic activity was associated with function recovery, indicating that Cr CEST MRI is promising to provide complementary information for heart remodeling at the molecular level.

Michael Bock¹, Serhat Ilbey¹, Alexander Maier², and Ali Caglar Özen^1
1Radiology - Medical Physics, University Medical Center, Faculty of Medicine, University of Freiburg, Freiburg, Germany, ²Cardiology and Angiology I, University Medical Center, Faculty of Medicine, University of Freiburg, Freiburg, Germany

Chemical shift artifact correction is achieved via spectral Hadamard encoding of the individual resonances using composite RF pulses. During decoding individual images of the resonances are created that can be combined to a final composite image. The method is demonstrated in ¹⁹F MRI of perfluorooctyl bromide that can be used to label monocytes for inflammation. Compared to other correction methods Hadamard decoding can be implemented directly in the image reconstruction, and the full SNR advantage of the image combination can be realized.  

Brianna F. Moon^1,2, Iris Y. Zhou^1,2, Yingying Ning^1,2, Sergey Shuvaev^1,2, Mariane M. Le Fur^1,2, Yin-Ching I. Chen¹, Eman A. Akam³, Jonah P. Weigand-Whittier¹, Nicholas Rotile^1,2, Matthew Drummond⁴, Avery T. Boice^1,2, Samantha E. Zygmont^1,2, Rod R. Warburton⁵, Barry L. Fanburg⁵, Krishna C. Penumatsa⁵, and Peter D. Caravan^1,2
1Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA, United States, ²Radiology, Institute for Innovation in Imaging, Massachusetts General Hospital, Charlestown, MA, United States, ³Medicine, Division of Cardiology, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA, United States, ⁴Medicine, Division of Pulmonary and Critical Care Medicine, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA, United States, ⁵Pulmonary, Critical Care and Sleep Medicine, Tufts Medical Center, Boston, MA, United States

Group 2 pulmonary hypertension is a complication of chronic left ventricular dysfunction. During the early stages of the disease activity there is an accumulation of extracellular matrix molecules and an increase in allysine content. Molecular MR imaging with an allysine-targeted fibrogenesis probe shows increased lung-to-muscle ΔCNR within the lungs of transverse aortic constriction (TAC)-induced left ventricular dysfunction animal models compared to control animals, which corresponded to increased lung and right ventricle weight, and elevation of fibrogenesis biomarkers (hydroxyproline and allysine).

Tobias Hoh¹, Valery Vishnevskiy¹, Christian T Stoeck^1,2, Conny F Waschkies², Nikola Cesarovic³, Miriam Weisskopf², Maximilian Fuetterer¹, and Sebastian Kozerke^1
1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland, ²Center of Surgical Research, University Hospital Zurich, University of Zurich, Zurich, Switzerland, ³Department of Health Sciences and Technology, ETH Zurich, Zurich, Switzerland

Three-dimensional first-pass myocardial perfusion CMR requires acceleration methods to enable whole-heart coverage in a limited acquisition window. Current 3D approaches suffer from data inconsistencies during free breathing, compromising image quality and perfusion quantification. We propose and experimentally validate a combination of Cartesian pseudo-spiral k-t undersampling with respiratory motion-informed locally low-rank (MI-LLR) reconstruction for robust whole-heart myocardial quantitative perfusion CMR in an experimental pig model of myocardial infarction. As reference, standard 2D acquisitions are used. Overall, image quality scoring was good. Perfusion defects were equally well depicted and discernible in imaging and MBF perfusion mapping in 3D and 2D approaches, respectively.

Carlos Velasco¹, Alina Hua¹, Anastasia Fotaki¹, Camila Munoz¹, Giorgia Milotta¹, Karl P. Kunze^1,2, Radhouene Neji^1,2, Tevfik F. Ismail¹, Claudia Prieto¹, and Rene M. Botnar^1
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, ²MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom

Myocardial tissue characterization including quantification of fibrosis and oedema plays an important role in the evaluation of many myocardial diseases. T₁ and T₂ maps are typically acquired sequentially in 2D under several breath-holds. However, misregistration artifacts, spatial resolution and volumetric coverage remain a limitation. To address this, a high resolution, 3D whole-heart motion-compensated joint T₁/T₂ water/fat sequence has been proposed and validated in a phantom and in healthy subjects. Here, we demonstrate the feasibility of the proposed approach to simultaneously acquire whole-heart co-registered T₁, T₂ maps as well as water and fat volumes in ~9min in patients with cardiovascular disease.

Jing Liu¹, Karen Ordovas², Yan Wang¹, Janine Lupo¹, Duan Xu¹, Yoojin Lee¹, Roselle Abraham¹, and David Saloner^1
1University of California San Francisco, San Francisco, CA, United States, ²University of Washington, Seattle, WA, United States

In this study, we developed a multi-compartment model based method for deriving multi-parametric mapping of the heart from the MOLLI T1 mapping acquisition. The preliminary results on patients with hypertrophic cardiomyopathy demonstrated the feasibility of deriving 9 quantitative parametric maps from a single scan and showed its potential of providing a comprehensive assessment of the myocardial tissue composition.

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

In the field of ultra-high field MRI, T₂* mapping is a promising technique for the non-invasive assessment of myocardial pathophysiology. Preclinical studies have already shown the potential to detect structural changes in infarcted myocardial tissue. Quantitative T₂* imaging is very demanding with regard to measurement-  and postprocessing time. Hence, we compared in this study the simpler and faster grayscale analysis of T₂*-weighted images with quantitative T₂* techniques. This was done in early and late acute infarct healing stages in a large animal model by comparing T₂* maps with grayscale T₂*-weighted images.

Marc Vornehm^1,2, Maximilian Fenski³, Elisabeth Preuhs⁴, Andreas Maier⁴, Jeanette Schulz-Menger³, Jens Wetzl¹, and Daniel Giese^1,5
1Magnetic Resonance, Siemens Healthcare GmbH, Erlangen, Germany, ²Artificial Intelligence in Biomedical Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, ³Working Group Cardiovascular Magnetic Resonance, Experimental and Clinical Research Center, Charité Medical Faculty, Max Delbrück Center for Molecular Medicine, HELIOS Klinikum Berlin Buch, Department of Cardiology and Nephrology, Berlin, Germany, ⁴Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, ⁵Institute of Radiology, University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany

Automatic segmentation of left ventricular myocardium on LGE images of patients with chronic myocardial infarction is challenging due to inhomogeneous signal intensities in the presence of myocardial scar. However, the inclusion of scar in automatically generated myocardium segmentations is critical for further LGE analysis. We propose a Deep Learning-based method for myocardial segmentation on cardiac LGE images, achieving average Dice scores of 0.78 and 0.80 on test images with and without scar, respectively. We furthermore evaluate the network’s sensitivity to scar and show that it can be improved by incorporating synthetic images with diverse enhancement patterns in the training data.