Abstract
ABSTRACT
The Iseult scanner delivered the first human brain images at 11.7 T. Here, we bring forth the design of the RF coil used for these acquisitions, featuring 8 transmit and 32 receive channels, then evaluate its SNR and g‐factor advantages with respect to a similar coil at 7 T. We attempt a deduction of the SNR dependency on field strength. First, we modeled our two compact coils for integration in a tight local shimming device. Electromagnetic simulation results on two different anatomical phantoms (one male and one female) were used for SAR/VOP, B1
+
, SNR‐ and g‐map predictions. Simulations were calibrated with acquisitions on a spherical agar phantom, and SNR measurements were based on a 3D‐GRE sequence. One volunteer was scanned at 7 and 11.7 T for that purpose, allowing a preliminary evaluation of in the ‐SNR model. Both coils performed as expected from simulations provided some corrections. For the agar sphere, we found a global intrinsic SNR gain of 2.75 between 7 and 11.7 T, corresponding to . In contrast the gain in the human brain was about 2.2, corresponding to , slightly less than previous results in the literature. Even though higher values are expected for an aqueous phantom because of its higher permittivity (increasing SNR thanks to the dielectric focusing effect), the main difference could be explained by physiological noise. As expected, the shorter wavelength induced by a higher also results in a perceptible but limited reduction of
g
‐factors in simulations and measurements. In conclusion, this study highlighted the benefits of a high in reception, without the potential bias of a different RF coil design; these benefits come with a slight decrease in transmission efficiency (due to additional losses in brain tissue at high frequencies) and a greater difficulty in homogenizing spin excitation.