Flow Quantification & Cardiovascular Image Processing
Click on to view the abstract pdf and click on to view the video presentation.
Friday May 13th
Room 512A-G  10:30 - 12:30 Moderators: Josh Oshinski and Smita Sampath

10:30 724.   Assessment of Left Ventricular 2D Pseudo Flow Pathway during Early Diastole Using SPAMM-PAV 
Ziheng Zhang1, Donald P. Dione2, Ben A. Lin2, Albert J. Sinusas2, and Smita Sampath1
1Department of Diagnositc Radiology, Yale University, School of Medicine, New Haven, CT, United States, 2Section of Cardiovascular Medicine, Yale University, School of Medicine, New Haven, CT, United States

Early diastolic filling is tightly coupled to regional myocardial relaxation patterns. Here, we track the 2D pseudo pathlines of blood emitter particles that are dynamically released from the mitral valve plane during early diastole. Results in normal volunteers demonstrate a distinct pattern of filling that corresponds closely with myocardial relaxation patterns. Further, results in an infarcted animal demonstrate significant abnormality, underlining the sensitivity of this method as an indicator of diastolic filling abnormalities.

10:42 725.   3D Aortic Blood Flow in Patients with Marfan Syndrome: Changes in Hemodynamics and Correlation with Aortic Geometry 
Michael Markl1, Julia Geiger1, Lena Herzer2, Brigitte Stiller2, and Raoul Arnold2
1Radiology, Medical Physics, University Medical Center, Freiburg, Germany, 2Pediatric Cardiology, University Medical Center, Freiburg, Germany

The aim of this study was to apply flow-sensitive 4D MRI for the comprehensive characterizations of flow pattern changes in clinically unsuspicious Marfan patients (without aortic disease) and to evaluate the effect of aortic geometry on hemodynamic alterations. Results were compared to findings in healthy volunteers and patients with suspected Marfan syndrome. Helix and vortex flow was clearly enhanced in the descending aorta of Marfan patients and was associated with significantly increased valve and aortic diameters. Flow-sensitive 4D MRI may help to identify Marfan patients at risk for cardiovascular complications and to assign appropriate therapies prior to these complications.

10:54 726.   Assessment of blood flow patterns in the Pulmonary Artery with 4D Flow MRI 
Pablo Bächler1, Natalia Pinochet1, Gérard Crelier2, Cristián Tejos3,4, Pablo Irarrazaval3,4, and Sergio Uribe4,5
1School of Medicine, Pontificia Universidad Católica, Santiago, Chile, 2Institute for Biomedical Engineering, University and ETH, Zurich, Switzerland, 3Electrical Engineering Department, Pontificia Universidad Católica, Chile, 4Biomedical Imaging Center, Pontificia Universidad Católica, Chile, 5Radiology Department, Pontificia Universidad Católica, Chile

4D Flow has been used to study blood flow patterns, mainly in the aorta. We performed a flow pattern analysis in the Pulmonary Artery (PA). 4D Flow depicted complex flow patterns in the PA, revealing two vortices in the main PA and one in the right PA in healthy subjects. Abnormal blood flow patterns were seen in patients with different Congenital Heart Diseases (CHD), revealing complex hemodynamics in the pulmonary circuit. Knowledge of abnormal flow patterns in the PA in patients with CHD has the potential to contribute to improve current surgical interventions in this group of patients.

11:06 727.   Absolute Quantification of Pulmonary Pressure Waveforms with MRI in Pulmonary Hypertension Patients 
Octavia Biris1,2, Sanjiv Shah3,4, Jeremy Collins1, Amir Davarpanah1, James Carr1,3, and Timothy J Carroll1,2
1Radiology, Northwestern University, Chicago, IL, United States, 2Biomedical Engineering, Northwestern University, Evanston, IL, United States, 3Feinberg School of Medicine, Northwestern University, Chicago, IL, United States, 4Cardiology, Northwestern University, Chicago, IL, United States

Pulmonary arterial hypertension, a disease of proliferation of the pulmonary arteries, is currently diagnosed by right heart catheterization. We propose to develop a non-invasive alternative to catheterization, by which pulmonary artery pressure is calculated according to the Windkessel model by direct convolution of the pulmonary artery flow wave with an exponential function of time and vessel parameters compliance and vascular resistance. While flow in the pulmonary artery branches can be measured directly by MRI, compliance must be estimated from measurements of flow and vessel cross-section throughout the cardiac cycle. Pulmonary vascular resistance is estimated from measures of right ventricular function.

11:18 728.   4D MR Velocity Mapping using PC VIPR to Quantify Blood Flow In Portal Hypertension 
Alejandro Roldán-Alzate1, Alex Frydrychowicz1, Eric J Niespodzany1, Benjamin R Landgraf1, Oliver Wieben1,2, and Scott B Reeder1,2
1Radiology, University of Wisconsin, Madison, WI, United States, 2Medical Physics, University of Wisconsin, Madison, WI, United States

Accurate quantification of hepatic blood flow would facilitate improved understanding of the hemodynamic mechanisms by which portal hypertension affects the function of the liver. In this retrospective analysis, data acquired with an efficient 4D phase velocity imaging method was analyzed to assess flow to the hepatic vasculature. Data from 17 participants, 10 with portal hypertension and 7 controls were evaluated. Compared to healthy controls, patients with portal hypertension demonstrated increased blood flow through the portal vein.

11:30 729.   4-dimensional magnetic resonance velocity mapping of blood flow patterns in chronic aortic dissections at 3T 
Alex Frydrychowicz1, Michael Markl2, Eric Niespodzany1, Christian Schlensak3, Mark Schiebler1, and Christopher J François1
1Department of Radiology, University of Wisconsin - Madison, Madison, WI, United States, 2Department of Radiology, Medical Physics, University Hospital Freiburg, Freiburg, Germany, 3Department of Cardiac Surgery, University Hospital Freiburg, Freiburg, Germany

The aim of this study was to evaluate the hemodynamics in patients with chronic aortic dissection using 4D velocity mapping. A large variety of altered flow patterns was observed associated with different extents of disease, vessel dilatation, and post therapeutic anatomy. In contrast to well-described normal aortic flow patterns, a multitude of additional vortices and helices was observed in these patients, indicating a high degree of potentially abnormal, disorganized flow. Changes in secondary flow patterns were also present, such as enhanced retrograde flow. This study demonstrates the feasibility of using 4D velocity mapping for the hemodynamic evaluation of chronic aortic dissection.

11:42 730.   Analysis of Reynolds, Strouhal and Womersley numbers in the healthy thoracic aorta 
Aurelien F Stalder1,2, Alex Frydrychowicz3, Max F Russe2, Jan G Korvink4,5, Jürgen Hennig2, Kun Cheng Li1, and Michael Markl2
1Dept. of Radiology, Xuanwu Hospital of Capital Medical University, Beijing, China, People's Republic of, 2Dept. of Radiology - Medical Physics, University Hospital Freiburg, Germany, 3Dept. of Radiology, University of Wisconsin, Madisson, United States, 4Dept. of Microsystems Engineering, University of Freiburg, Germany, 5Freiburg Institute for Advanced Studies (FRIAS), Freiburg, Germany

Turbulence and velocity fluctuations of the blood flow are believed to play a role in hemolysis, platelet activation and thrombus formation. Based on flow-sensitive MRI, Reynolds, Womersley & Strouhal numbers were measured in-vivo at 8 planes along the thoracic aorta in 30 healthy volunteers. The onset of turbulence (supra-critical Reynolds numbers) was evaluated based on a previous model describing the transition to turbulence for pulsatile flow. Onset of turbulence was most prominent in the ascending aorta and distal descending aorta. The results were shown to be statistically correlated with sex, body weight, aortic diameter and cardiac output.

11:54 731.   Accelerated Phase-Contrast MRI using Compressed Sensing and Parallel Imaging 
Daniel Kim1, Hadrien A Dyvorne1, Ricardo Otazo1, Daniel K Sodickson1, and Vivian S Lee1
1Radiology, Center for Biomedical Imaging, New York University School of Medicine, New York, New York, United States

Phase-contrast (PC) MRI is a promising modality for studying hemodynamics associated with pathophysiology, such as liver vascular flows in cirrhosis. A major disadvantage of PC MRI, however, is its low data acquisition efficiency, which may limit the achievable spatial and temporal resolutions within a clinically acceptable breath-hold duration. This study describes a method to accelerate PC MRI using compressed sensing and parallel imaging to jointly exploit image sparsity and coil sensitivity encoding. Seven healthy volunteers were imaged in hepatic and portal veins to validate accelerated PC MRI against PC MRI with GRAPPA. The resulting flow measurements were in good agreement.

12:06 732.   Pulse Wave Velocity Assessment in a Single Breathhold using Compartment k-t PCA 
Daniel Giese1,2, Tobias Schaeffter1, and Sebastian Kozerke1,2
1Division of Imaging Sciences and Biomedical Engineering, King's College London, London, United Kingdom, 2Institute for Biomedical Engineering, Univeristy and ETH Zurich, Zurich, Switzerland

Pulse wave velocity assessment using cine phase-contrast measurements suffers from long scan times due to the required high temporal resolution. We propose the acquisition of highly undersampled phase-contrast data in combination with compartment k-t PCA to enable acquisition of two flow encoded slices through ascending and descending aorta with a temporal resolution of under 10ms in a single breathhold. It is shown that the calculated pulse wave velocities compare well with literature values and show strong correlation with velocities measured using conventional fully sampled data acquisitions.

12:18 733.   Automated Cardiac Strain Estimation from 2D Cine DENSE MRI 
Andrew D. Gilliam1, Xiaodong Zhong2, Kenneth C Bilchick3, and Frederick H Epstein4
1Andrew D. Gilliam Consulting, Providence, RI, United States, 2MR R&D Collaborations, Siemens Healthcare, Atlanta, GA, United States, 3Cardiology, University of Virginia, Charlottesville, VA, United States, 4Radiology & Biomedical Engineering, University of Virginia, Charlottesville, VA, United States

Displacement encoding with stimulated echoes (DENSE) directly encodes tissue displacement into MR images, providing easy access to vital physiological information such as cardiac strain. Unfortunately, the quantification of displacement and strain from raw cine DENSE imagery currently relies on the manual delineation of cardiac anatomy. In this study, we present the first fully automated solution to estimate cardiac strain from 2D cine DENSE images. Results indicate good agreement between the innovative automated analysis algorithm and previously described methods.