Purpose

Flexible, easy-to-use high permittivity pads are currently used regularly in conjunction with state-of-the-art RF coils by many neuroscientists at 7T as a means of passive B1 shimming and/or SNR enhancement. While 7T systems are becoming more available to neuroscientists, a large portion of MR research is still accomplished at 3T. To enhance sensitivity at 125 MHz usually system upgrades are necessary. We are presenting a method to safely enhance currently available clinical systems by up to two-fold by simply replacing the foam cushion with our novel pad.

Methods

Using xFDTD (REMCOM, Inc., PA, USA) ten ultra-high dielectric constant (uHDC) material pad configurations with bulk permittivities between 600 and 1500 where numerically computed. The setup was similar to the in-vivo experiments with a 60cm diameter 2 CH quadrature driven birdcage coil and a 20 CH receive coil (Figure 1). The pad conformed to the visual cortex of Ella (Virtual Family, IT'IS Foundation, Switzerland). Transmit efficiency, SNR and 10g SAR (normalized B1+ in visual cortex) were assessed and based on these results a permittivity of 1200 for an 8 mm thick pad and diameter of 16 cm was selected. Using the series mixing rules a semiflexible circular composite pad with a bulk permittivity of 1200 was constructed from ceramics with a permittivity of 4700 and distilled water (HyQ Research Solutions, LLC, PA, USA). On a Siemens 3T Prisma we carried out the in-vivo experiments with and without this pad. On a 20-CH and the posterior portion of a 64-CH Siemens head coil (40-CH) each being standard clinical coils, transmit efficiency and SNR maps where acquired with the Bloch-Siegert technique. Using a stereo goggle setup, we presented an alternating flashing checker board paradigm to the eyes (10 Hz, 24 s on 24 s off) as shown in Figure 4. The goggles required the anterior portion of the 64 CH receive coil to be removed for all cases. During this visual stimulus paradigm we acquired high resolution EPI images (1x1x1.1 mm^3, TR=3s, TE=30ms) for fMRI of the visual cortex. The data was processed using SPM 12 (P<0.001 with extent threshold of 5 voxels).

Results and Discussion

The simulated results as shown in Figure 2 show that using this pad at a permittivity of 1200 while reducing both overall and peak 10g SAR. Both transmit efficiency and SNR were significantly improved. This was experimentally confirmed as shown in Figure 3 where transmit efficiency in the covered area was enhanced more than 2-fold. The SNR of the case using the pad in a20-CH head coil was improved 75-100% over its own baseline, to a similar level or up to 30% better level than the 64CH baseline. When using the composite pad in the 64-CH head coil, the SNR improved between 20-40% over the baseline. This was confirmed in the activation maps, which showed significant activation for the 20 and 64CH head coil using the pad but were not above the noise level without.

Conclusion

Our experimental and simulated results demonstrate that even though the highest SNR gains are likely to be achieved with monolithic ceramic materials, we are able to significantly improve SNR in standard clinical head coils at 3T. While more optimization work is needed to improve the quality of enhancement, with this flexible setup we already showed useful increases in transmit efficiency and SNR in a single subject fMRI test case. The benefits to the community could be substantial as it allows for significant SNR improvements of existing equipment without major technical investments.

Acknowledgements

Grants from the NIH and the Penn State Neuroscience Institute