View Presentation Video

Thekla Helene Oechtering^1,2, Tilman Schubert^1,3, Qianqian Zhao⁴, Roxana A Alexandridis⁴, Nikolaos Panagiotopoulos^1,2, Oliver Wieben^1,5, Kevin M Johnson^1,6, Alejandro Roldán-Alzate^7,8,9, and Scott B Reeder^1,5,8,9,10
1Department of Radiology, University of Wisconsin-Madison, Madison, USA, WI, United States, ²Department of Radiology, Universität zu Lübeck, Lübeck, Germany, ³Department of Neuroradiology, Zurich University Hospital, Zurich, Switzerland, ⁴Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, WI, United States, ⁵Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, United States, ⁶Department of Electrical and Computer Engineering, University of Wisconsin-Madison, Madison, WI, United States, ⁷Department of Radiology, University of Wisconsin-Madison, Madison, WI, United States, ⁸Department of Mechanical Engineering, University of Wisconsin-Madison, Madison, WI, United States, ⁹Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, United States, ¹⁰Department of Emergency Medicine, University of Wisconsin-Madison, Madison, WI, United States

Keywords: Liver, Velocity & Flow

Early prediction of remnant liver growth after portal vein embolization (PVE) would enable earlier surgery in patients with liver malignancies and thus decrease the risk of tumor progression. Portal blood flow after PVE holds the potential to be an important predictor for hypertrophy of the non-embolized segments. We demonstrated the feasibility of 4D flow MRI quantification of portal blood flow before and after PVE in a porcine model. Flow changes immediately after PVE were predictive of the change in liver volume 2 weeks post PVE in both the embolized and non-embolized liver lobes.

View Presentation Video

Hsueh-Ping Hung¹, Ming-Ting Wu², Ken-Pen Weng^3,4, and Hsu-Hsia Peng^1
1Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan, ²Department of Radiology, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, ³Department of Pediatrics, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, ⁴Department of Pediatrics, National Yang Ming Chiao Tung University, Taipei, Taiwan

Keywords: Heart, Cardiovascular

We aimed to evaluate the aortic blood pressure characteristics and cardiac function in patients after the Fontan operation. The Fontan group exhibited significantly increased peak diameter index and decreased peak velocity and peak flow rate in some of planes in the aorta. The Fontan group also presented decreased pressure difference (PD) in aortic arch and descending aorta. The aortic PD might be more sensitive than cardiac index in detection of cardiovascular alteration in Fontan patients. The aortic PD measured by 4D flow MRI can be a noninvasive alternative to characterize aortic remodeling.

View Presentation Video

Elizabeth Weiss¹, Justin Baraboo¹, Liliana Ma¹, Mariana B. L. Falcão², Christopher W. Roy², Joshua D. Robinson³, Matthias Stuber², Cynthia K Rigsby⁴, and Michael Markl^1
1Radiology, Northwestern University, Chicago, IL, United States, ²University of Lausanne (CHUV), Lausanne, Switzerland, ³Cardiology, Ann & Robert Lurie Children's Hospital, Chicago, IL, United States, ⁴Medical Imaging, Ann & Robert Lurie Children's Hospital, Chicago, IL, United States

Keywords: Flow, Velocity & Flow, congenital heart disease

We adapted our novel respiratory gating method to evaluate the impact of respiratory state (RS) resolution on respiratory driven flow measured by 5D flow MRI. We found that the impact of respiratory state resolution was both anatomy and vessel dependent. Caval veins and measurements in single ventricle disease patients were most impacted by the reduction of respiratory states. Shunt patients and pulmonary artery measurements were more robust to reduced RS and may be able to take advantage of the decreased acceleration associated with fewer RS. 5D flow MRI is well suited for this variable need as respiratory gating is retrospective.

View Presentation Video

Keywords: Flow, Simulations

Autonomous acquisition of radial k-space data improves sample uniformity in 2D cardiac MRI. We aim to extend this approach to 5D-flow imaging (4D flow with respiratory gating). We have simulated data acquisition using an autonomous approach (ARKS) and compared it to golden-angle (GA) based spiral phyllotaxis acquisition. Simulations were based on physiologic data recorded from pediatric patients undergoing conventional 4D flow imaging. We found that a 4.5 min ARKS scan achieves a higher degree of sampling uniformity than an 8 min GA scan. Future work will focus on implementing 5D-flow ARKS in vivo and evaluating the impact on flow accuracy.

View Presentation Video

Mariana B.L. Falcão¹, Giulia M.C. Rossi¹, Jonas Richiardi¹, Xavier Sieber¹, Pierre Monney², Tobias Rutz², Milan Prša³, Estelle Tenisch¹, Anna Giulia Pavon⁴, Panagiotis Antiochos², Matthias Stuber^1,5, and Christopher W. Roy^1
1Department of Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, ²Service of Cardiology, Centre de Resonance Magnétique Cardiaque (CRMC), Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, ³Woman- Mother- Child Department, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, ⁴Division of Cardiology, Cardiocentro Ticino Institute, Ente Ospedaliero Cantonale, Lugano, Switzerland, ⁵Center for Biomedical Imaging (CIBM), Lausanne, Switzerland

Keywords: Flow, Cardiovascular, Cardiac signal extraction

Self-gating (SG) techniques improve the ease-of-use of cardiac MR by deriving cardiac signals from the data itself, obviating the need for ECG lead placement. Nonetheless, unpredictable shifts between the features of SG signals and the conventionally used R-wave peaks from ECG might hamper a direct link of reconstructed image frames with physiology. In this work, we developed a fully convolutional neural network to predict R-wave peak timepoints from SG imaging readouts in free-running radial 4D flow data, and provided a proof-of-concept of the usability of such learned R-wave peak timepoints for reconstructing cardiac-resolved 4D flow images.

View Presentation Video

Pietro Dirix¹, Stefano Buoso¹, Valery Vishnevskiy¹, and Sebastian Kozerke^1
1University and ETH Zurich, Zürich, Switzerland

Keywords: Flow, Image Reconstruction

We investigate the impact of regularization on the quantification of flow and turbulence in highly undersampled data. To overcome the uncertainties due to the lack of ground truth, patient-specific aortic geometry and inflow conditions were extracted from in-vivo 2D cine and 2D phase-contrast MRI and used to generate synthetic personalized ground truth flow fields. Simulation results were embedded into the corresponding patient-specific 4D flow MRI effectively resulting in personalized synthetic datasets with known aortic ground truth and realistic background. The reconstruction of multiple undersampled datasets showed two distinct optimal regularization ranges for the quantification of flow velocities and turbulence.

View Presentation Video

Keywords: Flow, Data Acquisition

Tricuspid regurgitant velocity is a crucial biomarker in identifying pressure overload in the right heart, associated with diastolic dysfunction and pulmonary hypertension.  2D phase-contrast cannot quantify this flow, and echocardiography is used clinically. We developed a phase-contrast method which utilizes deep-learning algorithms to track the valvular slice in a cardiac phase-dependent manner, which we call 2.5D flow.  We studied its performance in nine healthy subjects and patients with tricuspid regurgitation. RV stroke volumes correlated better to forward flow volumes by 2.5D flow vs. static 2D phase-contrast (ICC=0.88 vs. 0.62).  2.5D flow characterized regurgitation in a patient.

View Presentation Video

Chiara Manini¹, Markus Hüllebrand^1,2, Marius Pullig³, Titus Kühne^1,4, Sarah Nordmeyer⁵, Lina Jarmatz⁵, Andreas Harloff⁶, Jeanette Schulz-Menger^4,5, and Anja Hennemuth^1,2,4,7
1Institute of Computer-assisted Cardiovascular Medicine, Charité – Universitätsmedizin Berlin, Berlin, Germany, ²Fraunhofer MEVIS, Bremen, Germany, ³IBM Germany, Berlin, Germany, ⁴DZHK (German Center for Cardiovascular Research), Partner site Berlin, Germany, ⁵Charité – Universitätsmedizin Berlin, Berlin, Germany, ⁶University Hospital Freiburg, Freiburg, Germany, ⁷University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Keywords: Flow, Velocity & Flow, Aorta segmentation

Standardized 4D PCMRI postprocessing protocols could enable comparable bloodflow quantification. We propose an automatic segmentation of aortic cross section over time with a residual trained data from different imaging sequences, scanner types, pathologies and position of cross section planes. Dice score, Hausdorff metric as well as flow and velocity curves for the segmented areas show good performance both in the validation and test sets.

View Presentation Video

Daan Bosshardt^1,2, Mitzi M. van Andel², Renske Merton¹, Vivian de Waard³, Roland R.J. van Kimmenade⁴, Arthur J. Scholte⁵, Moniek G.P.J. Cox⁶, Aeilko H. Zwinderman⁷, Aart J. Nederveen¹, Maarten Groenink², Pim van Ooij¹, and Eric M. Schrauben^1
1Radiology and Nuclear Medicine, Amsterdam UMC, Amsterdam, Netherlands, ²Cardiology, Amsterdam UMC, Amsterdam, Netherlands, ³Medical Biochemistry, Amsterdam UMC, Amsterdam, Netherlands, ⁴Cardiology, Radboud University Medical Center, Nijmegen, Netherlands, ⁵Cardiology, Leiden University Medical Center, Leiden, Netherlands, ⁶Cardiology, University Medical Center Groningen, Groningen, Netherlands, ⁷Clinical Epidemiology, Amsterdam UMC, Amsterdam, Netherlands

Keywords: Flow, Velocity & Flow, Pulse Wave Velocity

Pulse wave velocity (PWV) is a well-established measure to evaluate vessel wall stiffness and increases in the aorta of patients with Marfan Syndrome (MFS). With sufficient temporal resolution, 4D-flow MRI can be used for PWV calculations. In this study, we assess the repeatability of aortic PWV calculations in healthy volunteers using an open-source software tool and apply it in a MFS cohort. Our results show high repeatability of global aortic PWV, elevated PWV in MFS patients, and an association between elevated PWV and biomarkers for advanced disease.

View Presentation Video

Keywords: Data Processing, Velocity & Flow, 4D Flow MRI, Quantitative Hemodynamics

A generalized data-processing pipeline tool for performing completely automated hemodynamic assessment from raw 4D flow MR images is presented.  The tool is evaluated for performance in a group of 271 subjects with mixed distribution of healthy, valve disease, and connective-tissue disorder status.  A high success rate of 94% is achieved for fully-automated quantification of regional aortic peak velocities and global aortic pulse wave velocity.

View Presentation Video

Michael Loecher^1,2 and Daniel B Ennis^1,2
1Radiological Sciences Laboratory, Stanford University, Stanford, CA, United States, ²Radiology, Veterans Administration Health Care System, Palo Alto, CA, United States

Keywords: Flow, Velocity & Flow

Background phase errors in PC-MRI flow measurements are caused by eddy currents and mechanical vibrations that produce unwanted magnetic fields, thereby reducing velocity measurement accuracy.  In this work, we demonstrate that a GIRF measurement can be used to predict the PC-MRI background phase, then used to design gradient optimized (GrOpt) velocity encoding waveforms that minimize these errors.  The method is tested in static phantoms and in vivo, where 4.8x to 6.4x reductions in background velocity are seen.  Phantom results showed a reduction from 0.74±0.22 cm/s to 0.15±0.12 cm/s, and in vivo results showed reductions from 0.93±0.10 cm/s to 0.13±0.08 cm/s.

Wenjian Liu¹, Junpu Hu², Jiayu Zhu², Jian Xu³, Zijian Zhou¹, Haikun Qi^1,4, and Peng Hu^1,4
1School of Biomedical Engineering, ShanghaiTech University, Shanghai, China, ²United Imaging Healthcare, Shanghai, China, ³UIH America, Inc., Houston, TX, United States, ⁴Shanghai Clinical Research and Trial Center, ShanghaiTech University, Shanghai, China

Keywords: Flow, Velocity & Flow

Phase contrast MRI (PC MRI) has been widely used to quantify blood flow and velocity. Four-dimensional (4D) flow PC MRI needs to acquire the FC data and three-directional (3D) FE data interleaved within each cardiac k-space segment. In this work, we propose a more efficient flow encoding strategy for PC MRI using a temporal modulation technique and we showed the preliminary feasibility of quadrupling the temporal resolution or reducing the scan time by 70% compared with conventional 4D flow by redesigning and adjusting the temporal modulation strategy for under-sampled M1 space.

View Presentation Video

Johan Berglund^1
1Dept. of Surgical Sciences, Uppsala University, Uppsala, Sweden

Keywords: Flow, Velocity & Flow

The use of low velocity encoding (venc) improves the velocity-to-noise ratio in phase-contrast MRI, but increases the risk of phase wrapping. A novel phase-unwrapping technique based on graph-cuts was evaluated in five cardiac patients with high-, medium-, and low-venc acquisitions. The proposed method reduced the stroke volume error compared to a LaPlacian based reference method. In retrospectively wrapped data, the proposed method demonstrated excellent unwrapping results in most cases, even for venc as low as 25 cm/s. The graph-cut based method appears promising for low-venc acquisitions combined with phase-unwrapping.

View Presentation Video

Ioannis Koktzoglou^1,2 and Rong Huang^1
1Radiology, NorthShore University HealthSystem, Evanston, IL, United States, ²Pritzker School of Medicine, The University of Chicago, Chicago, IL, United States

Keywords: Flow, Brain, MRA

Quantitative time-of-flight (qTOF) magnetic resonance angiography (MRA) is a recently introduced technique that provides for simultaneous luminal and hemodynamic imaging of the intracranial arteries. We hypothesized that the application of a deep machine learning (DML) image analysis strategy to qTOF MRA data would improve agreement of intracranial arterial velocity measures with respect to phase contrast MRI.  Compared to a more conventional image analysis procedure, we found that the application of DML image analysis to qTOF data improved agreement of component, total, and peak intracranial arterial flow velocity measures with respect to phase contrast MRI, and reduced calculation times by 35-fold.

View Presentation Video

Pan LIU^1,2, Heimiri Monnier¹, Kimi Piedad Owashi¹, Serge Metanbou³, Cyrille Capel⁴, and Olivier Balédent^1,2
1CHIMERE UR 7516, Jules Verne University of Picardy, Amiens, France, ²Medical Image Processing Department, Amiens Picardy University Hospital, Amiens, France, ³Radiology Department, Amiens Picardy University Hospital, Amiens, France, ⁴Neurosurgery Department, Amiens Picardy University Hospital, Amiens, France

Keywords: Head & Neck/ENT, Velocity & Flow, real time phase contrast, phase contrast, respiratory effects, cerebral veins

Real-time phase-contrast sequences are increasingly used to study the effect of breathing on cerebral circulation, however, there is a lack of studies on the effect of free breathing (E-Fb) on cerebral veins. In this study, we quantified the intensity and phase shift of the E-Fb on cerebral veins in 19 healthy volunteers using a time-domain multiparameter analysis method. We analyzed the characteristics of the E-Fb on four parameters of four cerebral veins. We found that during expiration, the mean flow rate, amplitude, stroke volume, and cardiac period of internal jugulars veins and sinuses increased.