Session M1.2

Inhomogeneous Human Torso Model of Magnetohydrodynamic Blood Flow Potentials Generated in the MR Environment

GM Nijm*, S Swiryn, AC Larson, AV Sahakian

Northwestern University
Evanston, IL, USA

Magnetohydrodynamic (MHD) voltages resulting from blood flow in a magnetic field are superimposed on the ECG used for gating in the MR environment, which may result in triggering problems during MR image acquisition. Since MHD voltages are related to blood flow, characterization of the MHD signal may provide useful flow information. The equations which fully describe the generation of MHD voltages are Maxwell’s equations and Navier-Stokes equations.
Finite element methods may be used to numerically compute the solutions of these equations. We used Comsol Multiphysics modeling software to create a 3D model of the human torso in a 3.0 T static magnetic field. The model was derived from the dataset provided by the National Library of Medicine’s Visible Human Project. The structures represented in the model were the heart, aorta, lungs, ribs, sternum, spine, and skin. Physiologically realistic properties were used for the different tissue types.
The magnitude of the MHD voltages is related to the velocity of blood and the blood vessel diameter. We chose to model blood flow through the aorta since it has the highest velocity and largest diameter of all blood vessels. Accordingly, we expected the greatest contribution to the observed MHD voltages to arise from aortic blood flow. Blood velocity (m/s) was defined as: ((R^2–x^2–y^2)/R^2)*(3e^(-6t)*sin(2pt)), where R is the radius (m), t is time, and x and y are Cartesian coordinates from the vessel center (m). As such, the velocity had a spatially parabolic profile and varied in time. The model was solved for 50 ms time steps from 0 to 1.0s. The resulting MHD voltages were computed and modeled in 3D. These voltages were compared with experimentally acquired MHD voltages which were extracted from the ECG during cardiac MRI in nine normal subjects.
The model was solved with a mesh consisting of 30627 elements using an iterative solver (GMRES). The resulting model showed that the largest magnitude voltages arose in the transverse direction. The MHD voltage magnitudes varied as the velocity changed over time, with the maximum MHD voltage occurring at t = 0.15 s, and the minimum at t = 0.65 s, corresponding to the maximum and minimum input velocities, respectively. The experimental data had a mean maximum MHD voltage of 0.2 mV, compared with the modeled data which had a maximum of 0.285 mV.
In conclusion, by modeling MHD voltages in 3D, we can learn about their effect on the ECG during cardiac MRI, which may provide additional information about blood flow. In addition, by determining precisely where we expect the greatest MHD voltages to manifest on the surface, we can optimize the placement of ECG electrodes for cardiac MRI. Future work will include modeling flow through additional blood vessels to determine their contribution to the MHD voltages observed on the surface.

(Abstract Control Number: 145)