Introduction

Magnetic Resonance Imaging (MRI) is a non-invasive medical diagnostic technique, in which the nuclei are excited by the radio-frequency (RF) magnetic field known as the B1 field. For MRI systems that use a low static magnetic field (less than 0.5T), known as the B0 field, the Larmor frequency and hence the frequency of the B1 field is very low and the dimension of a human body is only a small fraction of the wavelength. In such a case, the interaction between the B1 field and the human body can be neglected and the B1 field can be evaluated in absence of the human body. Also, the electric field associated with the B1 field is negligible and so is the specific energy absorption rate (SAR).

However, because of the limitation of the signal-to-noise ratio (SNR) associated with low frequencies, low B0 field MRI systems cannot produce high enough SNR for some advanced studies such as functional MRI. To enhance the SNR, MRI systems using a high B0 field have been developed. However, as the strength of the B0 field increases, the frequency of the B1 field increases linearly. For example, the frequency of the B1 field for a 4T system is 171MHz for proton imaging. At such a high frequency, the interaction between the B! field and the human body can no longer be neglected. This interaction is caused by the dielectric resonance since the effective wavelength of the B1 field is now comparable to or even smaller than the dimension of the human body. Such a strong interaction not only degrades substantially the B1 field homogeneity and thus the imaging quality, but also can cause concerns about the safety because the electric field associated with the B1 field increases with the inhomogeneity of the B1 field.

In this work, we develop an efficient finite-element method (FEM) to investigate the SAR and the B1-field inhomogeneity of birdcage coils loaded with a human head. We choose birdcage coils in this study because they are the most popular ones used in MRI, mostly due to their capability to produce a highly homogeneous B1 field over a large volume within the coil.

Numerical Results

The electromagnetic model of a human head used in this study was constructed from a series of MR images and is shown in the following Figures 1a to 1c.

Figure 1a: Saggital slice through the head model

Figure 1b: Axial slice through the eyes of the head model

Figure 1c: Coronal slice through the head model

The SAR and B1 field distributions, excited by a linear and a quadrature birdcage coil, are calculated and presented at 64, 128, 171, and 256MHz. Figures 2 and 3 display the SAR (W/kg) of a loaded quadrature birdcage coil, and Figures 4 and 5 show its magnetic field (A/m) distribution.

Figure 2: SAR (W/kg) of a loaded birdcage coil in the axial slice (top left 64MHz, top right 128MHz, bottom left 171MHz, bottom right 256MHz)

Figure 3: SAR (W/kg) of a loaded birdcage coil in the saggital slice (top left 64MHz, top right 128MHz, bottom left 171MHz, bottom right 256MHz)

Figure 4: Magnetic field (A/m) of a loaded birdcage coil in the axial slice (top left 64MHz, top right 128MHz, bottom left 171MHz, bottom right 256MHz)

Figure 5: Magnetic field (A/m) of a loaded birdcage coil in the saggital slice (top left 64MHz, top right 128MHz, bottom left 171MHz, bottom right 256MHz)

The method presented here can be used to evaluate current and new designs of MRI RF birdcage coils. Accurate information about the B1 field can be used for MR spectroscopy and for designing new imaging schemes that compensate for the field inhomogeneity. It is also useful for a better understanding of the EM-NMR transduction in human anatomy. Accurate information about the SAR can be used to assess more accurately the potential hazards of RF fields on the patient.

The above work is a collaboration between Dr. Ji Chen and Prof. Jianming Jin. Please send suggestions, comments, and inquiries to: j-jin@uiuc.edu.