Acquiring homogeneous images with adequate flip angle is a challenge at ultra high field (UHF, >=7T) due to RF effects including shallow penetration, short wavelengths, destructive interferences, and increased SAR. Traditionally, transmit (B1+) and receive (B1-) arrays are composed of relatively narrow conductive elements aligned in the longitudinal direction, resulting in low homogeneity in the transverse plane for each element, requiring multiple transmit elements to obtain appreciable coverage on the transverse plane. We introduce a novel coil for high-fidelity axial imaging using a transverse-oriented electrically-long slot in a conductor, or “slot antenna,” such that a single coil can replace multiple longitudinal elements for production of a homogeneous transverse image (Fig. 1). According to Babinet’s Principle [1], if the slot and dipole are oriented in the same direction, they have similar radiation patterns, but orthogonal polarizations. Thus, a transverse slot antenna will produce the desired field polarization for MRI. Here with simulations and experiments we demonstrate, by many measures, performance of a single slot antenna is advantageous compared to a single dipole.

Simulations

A 0.8cm x 0.4cm x 0.8cm rectangular phantom with relative permittivity of 64.5 and conductivity of 0.62 S/m, was modeled using the Comsol Multiphysics 4.4 finite element modeling (FEM) solver (Burlington, MA, USA). Dipole and slot antennas were modeled 1cm above the phantom. The dipole antenna had a width of 1cm and length of 50cm (λ/2). For the slot antenna, conductor of length 52cm x 25cm was modeled with a 1cm by 50 cm slot in the middle. Both antennas were driven with a 1V source across the middle of their respective gaps (Fig 1A). The field distributions for both antennas were calculated at 300 MHz. For comparison of B1+, the fields of both coils were normalized by the square root of the maximum 10g average SAR (SAR10g). For comparison of SAR10g distribution, the voltage in each case was adjusted to produce an average B1+ on the transverse plane of 0.01mT. The normalized B1+ and SAR10g were plotted on axial and sagittal planes through the center of the phantom and a coronal slice 4cm inside the phantom.

Experiments

The slot and dipole antennas were constructed from PCB boards and matched to <-15dB when positioned next to the phantom/body. The dipole and slot antennas were placed independently 1cm from a body phantom (relative permittivity of 64.5, conductivity of 0.62S/m). Before the SNR measurements, B1+ maps were used to determine the reference voltage producing a 90-degree flip at 2cm into the phantom. The reference voltage was 145V for both antennas. High resolution spoiled axial GRE images were acquired with 10° flip angle, using TE=10ms, TR=1000, matrix size of 128x128 and FOV of 300mmx300mm for axial and sagittal slices at through center of the phantom and coronal slice 4cm inside the phantom. Noise data were acquired with zero transmit voltage. Flip angle maps of the GRE acquisitions were obtained using pre-saturation based flip angle mapping [2]. Finally, the slot antenna was placed above the hips of a volunteer and GRE images were acquired in the three orthogonal planes.

EM field simulation results demonstrate superior B1+ coverage and homogeneity on transverse and coronal planes using the slot antenna for a given max 10g SAR (Fig. 2B), making the slot antenna highly effective for UHF imaging within given SAR limits. Maximum 10g average SAR to produce 0.01mT average B1+ on the transverse plane was roughly ½ that of the dipole antenna (Fig. 2C). In phantom experiments, the slot antenna produced superior coverage and penetration (Fig 3B&C) with ~35% greater SNR and 45% greater flip angle at a location 4cm into the phantom. In vivo images of the pelvis (Fig 4) demonstrate excellent coverage and penetration on all three orthogonal planes using a single slot antenna.We have demonstrated a novel slot antenna for improved coverage and penetration on the transverse plane. Results indicate much greater coverage and homogeneity for a slot antenna than for a dipole on transverse and coronal planes for a given max SAR10g, and (due to their more distributed E fields) much lower maximum local SAR for the slot for a given average B1+. In vivo images show excellent coverage of the entire anterior portion of the pelvis using a single coil. Slot antennas are potentially very valuable in the quest for high homogeneity and low SAR at high field.