Matthew J. Middione¹, Julio A. Oscanoa^1,2, Ali B. Syed¹, Shreyas S. Vasanawala¹, and Daniel B. Ennis^1,3
1Department of Radiology, Stanford University, Stanford, CA, United States, ²Department of Bioengineering, Stanford University, Stanford, CA, United States, ³Cardiovascular Institute, Stanford University, Stanford, CA, United States

We have previously demonstrated a Deep Learning (DL) reconstruction of highly accelerated 2D PC-MRI datasets. We trained a DL reconstruction by retrospectively undersampling fully-sampled 2D PC-MRI datasets to enable up to 9x accelerated images with <5% error in accuracy and precision. Herein, we compare the accuracy and precision of 2D PC-MRI measurements for our prospectively deployed sequence and DL reconstruction. In this initial feasibility study, we show that our DL reconstruction provides <5% error in the accuracy of peak velocity and total flow relative to 2x parallel imaging and better accuracy and precision compared to 8x compressed sensing.

Christopher W Roy¹, Jérôme Yerly^1,2, Milan Prša³, Estelle Tenisch¹, Tobias Rutz⁴, Davide Piccini^1,5, and Matthias Stuber^1,2
1Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, ²CIBM Center for Biomedical Imaging, Lausanne, Switzerland, ³Woman-Mother-Child, Pediatric Cardiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, ⁴Cardiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, ⁵Advanced Clinical Imaging Technology (ACIT), Siemens Healthcare AG, Lausanne, Switzerland

A novel methodology for whole-heart 4D MRI with retrospective compensation of both respiratory motion and bulk patient movement is developed and its initial feasibility demonstrated. This approach enables imaging with high isotropic resolution and allows for retrospective evaluation of the dynamic cardiac anatomy creating a new way to significantly improve image quality in un-cooperative patients and potentially decreasing the need for sedation.

Olivier Jaubert^1,2, Javier Montalt-Tordera¹, Simon Arridge², Jennifer Steeden¹, and Vivek Muthurangu^1
1Institute of Cardiovascular Sciences, University College London, London, United Kingdom, ²Department of Computer Science, University College London, London, United Kingdom

Real-time cardiac output monitoring is desirable but requires low latency reconstruction and segmentation. A framework based on a spiral real-time flow acquisition and joint deep artifact suppression and segmentation was developed to provide flow volumes with low latency at the scanner. The proposed network was characterized in simulations and the full framework tested at the scanner providing real-time monitoring of flow during exercise. Future work will aim to validate the method in a larger cohort.

Thekla Helene Oechtering^1,2, A M K Muntasir Shamim³, Alma Spahic⁴, Nikolaos Panagiotopoulos^1,2, Oliver Wieben^1,4, Alejandro Roldán-Alzate^1,5,6, Scott B Reeder^1,4,5,6,7, and Kevin M Johnson^1,4
1Department of Radiology, University of Wisconsin-Madison, Madison, WI, United States, ²Department of Radiology and Nuclear Medicine, Universität zu Lübeck, Lübeck, Germany, ³Department of Electrical and Computer Engineering, University of Wisconsin-Madison, Madison, WI, United States, ⁴Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, United States, ⁵Department of Mechanical Engineering, University of Wisconsin-Madison, Madison, WI, United States, ⁶Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, United States, ⁷Department of Emergency Medicine, University of Wisconsin-Madison, Madison, WI, United States

4D flow MRI of the abdomen is challenging in obese patients due to confounding streak artifacts from fat and large imaging volume. This ongoing study aims to test different fat mitigation strategies to improve image quality. We acquired three 10-min contrast-enhanced, radial 4D flow sequences: 1) with clinically used standard parameters, 2) with inner volume excitation, and 3) with fat saturation. Preliminary results confirm best signal-to-noise ratio, longest streamlines and fewest streak-artifacts using inner volume excitation. Flow rate and repeatability did not differ between sequences.

Carola Fischer^1,2, Jens Wetzl², Tobias Schäffter^1,3,4, and Daniel Giese^2,5
1Department of Medical Imaging, Technical University of Berlin, Berlin, Germany, ²Magnetic Resonance, Siemens Healthcare GmbH, Erlangen, Germany, ³Physikalisch-Technische-Bundesanstalt (PTB), Braunschweig and Berlin, Germany, ⁴School of Imaging Sciences and Biomedical Engineering, King's College London, London, SE1 7EH, United Kingdom, ⁵Institute of Radiology, University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Erlangen, Germany

4D Flow data requires background phase removal to correctly quantify flow velocities. Typical methods are semi-automatic, requiring manual parameter input and remain time-consuming. Furthermore, correction methods based on stationary tissue segmentation fail when in the presence of wrap-around artifacts. In contrast, the proposed M-Estimate SAmple Consensus (MSAC) algorithm robustly and fully automatically corrects background phases by rejecting pixels that show large deviations from the most probable phase offset fit. A parameter sensitivity analysis is proposed following in vivo application in 13 datasets.

Thara Nallamothu^1,2, Haben Berhane^1,2, Liliana Ma¹, Justin Baraboo^1,2, Daniel C. Lee³, Daniel Kim¹, Phillip Greenland⁴, Michael Markl^1,2, Rod Passman³, and Mohammed S.M. Elbaz^1
1Radiology, Northwestern University, Chicago, IL, United States, ²Biomedical Engineering, Northwestern University, Chicago, IL, United States, ³Medicine (Cardiology), Northwestern University, Chicago, IL, United States, ⁴Preventative Medicine, Northwestern University, Chicago, IL, United States

Eddy current corrections are essential in preprocessing 4D flow data due to differing offset errors in each flow component leading to spatiotemporal inaccuracies in the velocity vector-field. Current automated optimization methods may require training data, have limited generalizability to different scan protocols, are sensitive to wraparound errors, and do not account for complex spatiotemporal errors. We introduce a self-calibrating method to optimize eddy current corrections and detect wraparound errors by using our recent 4D flow-field disparity signature technique that stochastically encodes the entire spatiotemporal profile of paired vector disparities and show feasibility and generalizability to two different 4D flow protocols.

Seraina Anne Dual^1,2, Mattia Arduini¹, Brigitte Schmittlein ³, Judith Zimmermann^1,4, Ellen Roche⁵, and Daniel Ennis^2,6,7
1Department of Radiology, Stanford University, Palo Alto, CA, United States, ²Cardiovascular Institute, Stanford University, Palo Alto, CA, United States, ³Department of Mechanical Engineering, Stanford University, Palo Alto, CA, United States, ⁴Department of Computer Science, Technical University of Munich, Garching, Germany, ⁵Institute for Medical Engineering & Science, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁶Department of Radiology, Stanfor University, Palo Alto, CA, United States, ⁷Division of Radiology, Veterans Affairs Health Care System, Palo Alto, CA, United States

MRI can be used to evaluate the interaction of soft robotic devices with biological tissues. The aim of this study was to use MRI to optimize the actuation pressure of an extra-aortic soft robotic cardiac assist device. We evaluated wall deformations, displaced fluid volumes and qualitative flows induced by the actuator on a fluid filled elastic vessel. Our results show that the displaced volume increases with the actuator pressure and decreases with increasing the vessel pressure. The maximum actuator pressures do not lead to buckling of the vessel wall given the current three-bladder design.

Cecilia Björnfot¹, Anders Eklund^1,2, Jenny Larsson³, William Hansson³, Sara Qvarlander¹, Lars-Owe Koskinen³, Jan Malm³, and Anders Wåhlin^1,2,4
1Department of Radiation Sciences, Umeå University, Umeå, Sweden, ²Umeå Center for Functional Brain Imaging (UFBI), Umeå University, Umeå, Sweden, ³Department of Clinical Science, Neurosciences, Umeå University, Umeå, Sweden, ⁴Department of Applied Physics and Electronics, Umeå University, Umeå, Sweden

Arterial stiffening occurs with age and could be detrimental to the brain through several pathways. This potentially aggravates factors causing cerebral small vessel disease. We assessed arterial stiffness of the intracranial arteries through a newly developed global intracranial pulse wave velocity (PWV) measurement using 4D flow MRI data in a population-based cohort of older people and compared it to white matter hypointensity volume (WMH). Results revealed a weak but significant correlation between WMH and PWV, suggesting that such PWV measurement could be a useful tool to assess the brain’s vascular health. 

Ashifa Hudani¹, David Patton ², James A White³, Steven Greenway², and Julio Garcia^3
1University of Calgary, Calgary, AB, Canada, ²Alberta Children's Hospital Research Institute, Calgary, AB, Canada, ³Libin Cardiovascular Institute, Calgary, AB, Canada

Tetralogy of Fallot (TOF) occurs in 1 in 3,500 births and is the most common cyanotic congenital heart defect. Patients with repaired TOF (rTOF) require constant monitoring to prevent life-threatening adverse effects. Hence, this study used 4D Flow MRI to assess pulmonary flow hemodynamics in patients with rTOF. All hemodynamic parameters calculated along the main pulmonary were shown to be higher in patients compared to controls. 

Maurice Pradella^1,2, Justin J Baraboo¹, Suvai Gunasekaran¹, Mitchell A Collins¹, Anthony Maroun¹, Amanda L DiCarlo¹, Rishi Arora³, Phil Greenland⁴, Rod Passman³, and Michael Markl^1
1Department of Radiology, Northwestern University, Chicago, IL, United States, ²Department of Radiology, University Hospital Basel, University of Basel, Basel, Switzerland, ³Department of Cardiology, Northwestern University, Chicago, IL, United States, ⁴Department of Preventive Medicine, Northwestern University, Chicago, IL, United States

Atrial fibrillation (AF) significantly increases stroke risk which is attributed to thrombus formation in the left atrium (LA) and to a greater extent the LA appendage (LAA). In this study, we investigated 4D-flow MRI-derived blood flow metrics from pre- and post-ablation exams in 26 AF patients undergoing pulmonary vein isolation (PVI). Blood stasis in both the LA and LAA significantly decreased in the post-PVI 4D-flow MRI. Furthermore, LA and LAA volumes also significantly decreased. Our results suggest that 4D-flow MRI was sensitive to detect treatment-related changes of LA and LAA flow metrics implicated on thromboembolic risk in AF patients.   

James Rice^1,2, Labib Shahid^1,2, Haben Berhane^3,4, Cynthia K Rigsby⁵, Joshua D Robinson⁵, Lindsay M Griffin⁵, Michael Markl^3,4, and Alejandro Roldan-Alzate^1,2
1Mechanical Engineering, University of Wisconsin-Madison, Madison, WI, United States, ²Radiology, University of Wisconsin-Madison, Madison, WI, United States, ³Radiology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States, ⁴Biomedical Engineering, Northwestern University, Evanston, IL, United States, ⁵Medical Imaging, Lurie Children's Hospital of Chicago, Chicago, IL, United States

4D flow MRI can be used to assess aortic coarctation (COA) repair hemodynamics in-vivo, however, challenges arise depending on hemodynamics and clinical presentation. This study presents a modeling pipeline for in-vitro COA hemodynamic assessment. Three patient-specific in-vitro models were created representing pre- and post-repair COA geometries, connected to a pulsatile flow loop, and imaged with 4D flow MRI to assess model hemodynamics. In all cases, aortic flow profiles are shown to improve after intervention. This agrees with clinical expectation, suggesting in-vitro 4D flow MRI can be used to analyze COA hemodynamics.

Shirin Aliabadi¹, Michael S. Bristow¹, Carmen Lydell¹, Paul WM Fedak¹, James White¹, and Julio Garcia^1
1University of Calgary, Calgary, AB, Canada

We aimed to detect the reliability of flow component analysis and intra-cardiac pressure drop analysis in the bicuspid aortic valve (BAV)-induced regurgitation cases compared to age-matched healthy control in the preserved ejection fraction (EF). Flow component analysis did not show differences while intra-cardiac pressure drop differentiated groups. 

Camila Munoz¹, Alina Hua¹, Karl P Kunze², Radhouene Neji^1,2, Tevfik F Ismail¹, René M Botnar¹, and Claudia Prieto^1
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, ²MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom

Dark-blood imaging is a promising tool for vessel wall assessment in cardiovascular disease. Here we introduce a novel water/fat DANTE-prepared sequence for 3D dark-blood imaging of the thoracic and abdominal aorta. The sequence integrates image navigation to enable respiratory motion correction resulting in a predictable scan time. Results from healthy subjects show the feasibility of the sequence for depicting the aortic wall with high-resolution with an overall short scan of ~6 minutes.

Shubhajit Paul¹, Raphael Tomi-Tricot^1,2, Camila Munoz¹, Karl Philipp Kunze^1,2, Shaihan Malik^1,3, Joseph V. Hajnal^1,3, Philippa Bridgen^1,3, Thomas Wilkinson^1,3, Sharon Giles^1,3, Radhouene Neji^1,2, Reza Razavi¹, Claudia Prieto^1,4, and René Michael Botnar^1,4
1School of Biomedical Engineering and Imaging Sciences, King’s College London, London, United Kingdom, ²MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom, ³The London Collaborative Ultra high field System (LoCUS), London, United Kingdom, ⁴Escuela de Ingeniería, Pontificia Universidad Católica de Chile, Santiago, Chile

3D cardiovascular MR imaging at ultra-high field (7T) is very challenging due to magnetic-field inhomogeneity (B₀ and B₁), SAR constraints and also the coil-depth issues of surface coils. To our best knowledge this is the first report demonstrating contrast-free ultra-high resolution (0.7mm³ isotropic) free-breathing 3D CMRA images with translational respiratory motion compensation. To avoid SAR constraints imposed by the use of T₂-preparation pulses, high excitation flip-angles are exploited to obtain high contrast between the coronary arteries and myocardium. This technique may be adopted for clinical applications to produce high-resolution cardiac MR images at 7T scanners.

Eva Peper¹, Hannes Dillinger¹, Charles McGrath¹, Christian Guenthner¹, and Sebastian Kozerke^1
1Institute for Biomedical Engineering, ETH Zurich, Zurich, Switzerland

The use of balanced steady-state free precession sequences (bSSFP) is an approach to compensate for the signal-to-noise ratio (SNR) loss at lower-field systems (<1T). Low field MRI systems are cheaper, lighter, and less prone to system imperfections than regular 1.5-3T systems, used in clinical practice. In this work, the feasibility and accuracy of velocity encoding using phase-contrast balanced SSFP (PC-bSSFP) is demonstrated and tested on a clinical 3T MRI system ramped down to 0.75T. Phantom and in-vivo results are compared to a standard 2D spoiled gradient echo phase-contrast sequence (PC-GRE).