Analysis of a Prototype MRI Hybrid Birdcage RF Coil with Uncertainty Quantification (*)

J. T. Fong [1],
[1] National Institute of Standards & Technology, Gaithersbug, MD, USA
Published in 2016

  1. INTRODUCTION. In a magnetic resonance imaging (MRI) system (see Fig. 1), it is necessary to excite the nuclei of a patient into coherent precession for imaging. This requires coupling between the nuclei and a source of radio frequency (RF) power (the transmitter). To receive a meaningful signal, one also needs to couple the nuclei to an external circuitry (the receiver). These two devices are called RF coils or resonators, and are key components in an MRI system (see Fig. 2, after Ref. [1]).

  2. USE OF COMSOL MULTIPHYSICS® SOFTWARE. In this paper, we present the finite-element-method-based solutions at infinite degrees of freedom, with uncertainty quantification, of the first resonance of a prototype MRI hybrid 16-legged and 2-port birdcage coil (see Fig. 3) using the RF Module of COMSOL Multiphysics® software. To quantify the uncertainty due to element type, we used two types of 3D mesh operations in COMSOL® software, namely, (a) the free tetrahedrons, and (b) the swept type involving layers of prism or hexahedral elements. To quantify the uncertainty due to mesh density, we applied a nonlinear least squares 4-parameter logistic function fit technique (see Ref. [2]) to five candidate solutions of increasing degrees of freedom such that a solution at infinite degrees of freedom can be estimated.

  3. RESULTS. In designing the hybrid birdcage for the finite element method-based analysis, we introduced free parameters not only for the 16 capacitances in the upper and lower rungs, but also capacitances in the 16 legs of the coil. We also introduce a parameter for refining the mesh density. The results of the analysis, when expressed in the absolute value of S11, the power parameter for the reflected wave of the active port of the birdcage coil, are shown in Fig. 4 for the two types of elements we used, namely, the all tetrahedron-runs and the swept type (mixed prism-hexahedrons). The extrapolated results of the two element types, 3.82 dB for the all tetrahedron type, vs. 4.39 dB for the swept type, differ by about 13 %.

  4. CONCLUSION. We show in this paper that element type and mesh density play a significant role in determining the accuracy and uncertainty of a finite-element-method-based solution of a typical problem in electromagnetics.

(*) Contribution of the U.S. National Institute of Standards & Technology. Not subject to copyright.