Control of Excited Modes in Ultra High Magnetic Field MRI with Electrically Hard Surfaces
Patrick Bluem1, Andrew Kiruluta2, Pierre-Francois Van de Moortele3, Gregor Adriany3, and Zoya Popovic1

1Department of Electrical, Computer, and Energy Engineering, University of Colorado at Boulder, Boulder, CO, United States, 2Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States, 3University of Minnesota, Center for Magnetic Resonance Research, Minneapolis, MN, United States

Synopsis

Traditional MRI reactive near-field probe design for B1 field uniformity assumes quasi-static fields. However, for B0>4T, the quasi-static approximation is no longer valid since the wavelength is smaller than the FOV and field wave modes appear, affecting image quality. This work presents the use of a copper strip waveguide structure combined with a traveling wave excitation at 7T, 10.5T human wide-bore and 16.4T small animal scanners, while observing the effect on a cylindrical distilled water phantom. A simple flexible copper strip wearable wrap is shown to improve SNR and field distribution in UHF-MRI.

Purpose

To improve field coupling and control using a hard surface waveguide that increases the SNR and B1 field homogeneity in a homogeneous phantom imaged in UHF-MR systems (7T and above).

Methods

Traditional MRI reactive near-field probe design for B1 field uniformity assumes quasi-static fields. However, for B0>4T, the quasi-static approximation is no longer valid since the wavelength is smaller than the FOV .1 The increase in frequency results in the formation of field modes with high field non-uniformity, undesirable in MRI. One solution to the problem is an electrically “hard surface” (EHS) flexible cylindrical waveguide wrapped around the imaging volume. Such an EHS is formed of a longitudinally corrugated surface, where the tangential electric and longitudinal magnetic fields are equal to zero along the surface.2 Fundamentally, the EHS modifies the boundary conditions of the imaging volume, thus changing the field profile. A simple EHS is implemented with copper tape on a thin flexible plastic sheet for use at 7T, 10.5T human wide-bore and 16.4T small animal scanners. At 7T and 10.5T, 3cm-wide copper strip spacing is 3 cm and the entire structure is wrapped around a distilled water cylindrical phantom (L=37.5cm, D=16cm). At 16.4T, 3cm-wide copper strips are spaced 1cm apart, due to both a smaller wavelength and phantom diameter (L=34 cm, D=9cm).

Results

A series of experiments were performed with right-handed circular polarized patch travelling wave (TW) probe excitations (Fig. 1) for all three systems (Fig. 2a). The circular patches, scaled to the respective Larmor frequencies, were placed 15 cm from the phantom for the 7T and 10.5T experiments. This was experimentally found to be the optimal probe distance which did not produce an overpowering near-field contribution that degrades the circular polarization, thus losing power to the cross-polarized field modes. The reference voltage of the excitation is kept constant across experiments on the same scanner. The resulting image without the EHS (Fig. 3a) has relatively low SNR and poor field uniformity with only one probe excitation. For the first EHS experiment, a 17-cm long EHS was placed around the air-gap between the probe and the phantom in the 7T bore, placed 1cm away from the probe with 3cm of the EHS wrapped around the phantom, greatly increasing the field coupled to the proximal side, but with the field attenuating rapidly along the phantom length (Fig. 3b). Next, only the full 37.5-cm length of the phantom is covered with the EHS, creating two distinct high field regions, with greatly reduced field in the center (Fig. 3c). The higher field strength farthest from the probe is due to an end reflection and would not be present in a material with higher loss or a large imaging volume. However, the relative magnitude of the field is much lower than in the previous case. The final 7-T experiment has both the air gap and the phantom surrounded by one complete EHS wrap, resulting in a higher SNR and overall better quality image (Fig. 3d). The overall benefit from the EHS observed in simulations is the 5 times larger B1 strength relative to the case with no EHS. The overall change is also apparent at 10.5T and 16.4T. At 10.5T, the probe was placed 15cm away from the phantom and both the air gap and the complete phantom were covered in the EHS, resulting in a drastic increase in the relative field magnitude (Fig.4), with a focused field distribution at the end closest to the exciter probe, while the center of the phantom is almost invisible. At 16.4T, the setup is fairly different (Fig. 2b) with a large discontinuity between the bore and the gradient coil region. For this system, the EHS not only increased the SNR seven times (Fig. 5) compared to an image with no guiding structure, but also yielded transmit B1 magnitude large enough to apply a refocusing RF pulse and collect spin echo images (not shown) with the patch probe excitation scaled to 698MHz.3

Conclusions

A simple flexible wearable copper strip electrically-hard surface (EHS) structure is shown to improve SNR and field distribution in UHF-MRI. With more complicated periodic flexible structures, boundary conditions of the electromagnetic field on the phantom can be tailored for a desired field distribution. This flexible EHS wrap can also be used with classical MRI volume coils, as the boundary conditions on the phantom do not depend on the excitation.

Acknowledgements

The authors acknowledge support by the National Science Foundation under a collaborative research grant ECCS 1307614.

References

1. D. O. Brunner, et al. “Travelling-wave nuclear magnetic resonance,” Nature 457(7232). 994-U2. 2009.

2. G. Ruvio, et al. "Modal propagation in ideal soft and hard waveguides," IEEE APS Intern. Symp Dig., vol.4, no., pp.438-441 vol.4, 22-27 June 2003.

3. P. Bluem, et al. "Travelling-wave excitation for 16.4T small-bore MRI," IEEE MTT-S Intern. Symp. Dig. pp.1-4, 17-22 May 2015

Figures

Layout of 7T (297 MHz) circularly-polarized patch probe. The same type of probe is scaled to 447 MHz for 10.5T and 698MHz for 16.4T scanners.

(a) Dimensions of the 7T and 10.5T scanners in the gradient coil region (additional bore length not shown) (b) Scanner dimensions for 16.4T. (c) Position of EHS showing the 15-cm gap wrapped (1), and 37.5cm phantom wrapped (2). The last experiment combines (1) and (2) for full coverage.

Measured 7T images, normalized and plotted on a linear scale, for: (a) no EHS; (b) case (1) from Fig.2c; (c) case (2) from Fig.2c; and (d) cases (1) and (2) combined, with EHS wrapped around both gap and phantom.

Measured 10.5T images for: (a) no EHS and (b) cases (1) and (2) from Fig.2c combined, with EHS wrapped around gap and phantom.

Measured 16.4T images for: (a) no EHS and (b) EHS covering both gap and phantom.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
3534