Synopsis

In this abstract we propose a new antenna design as a transmit array element for proton head imaging at 7T that combines a small feeding dipole that is capacitively coupled, as a so-called ‘indirect’ or ‘passive’ feed, to an ‘Egyptian’ axe dipole. B₁⁺ and SAR efficiency of the proposed antenna have been compared numerically with a conventional dipole antenna of the same length. A head array consisting of eight antenna elements has been constructed and tested. This array element offers a large reduction in SAR compared to equally sized linear dipoles while maintaining their B₁⁺ profile as well as reducing the impact of antenna loading on the resonance frequency.

Introduction

Neuroimaging at 7 Tesla is complicated by interferences that occur in the B₁⁺ field as a result of the relatively short wavelength of the RF transmit field compared to the dimensions of the human head¹. RF shimming by controlling the relative phases of individual transmit elements in a transmit array allows for an increased homogeneity of the transmit field in the human head compared to a standard circularly polarized excitation^2,3. With the increased availability of multi-channel transmit systems, many array designs for neuroimaging in UHF-MRI have been proposed^2,4,5.

In this abstract we propose a new design for a dipole antenna array for proton head imaging at 7T that combines a small feeding dipole that is capacitively coupled, as a so-called indirect or passive feed, to an ‘Egyptian’ axe dipole (EAD)⁶.

Methods

The EAD antenna was printed on a 0.5 mm thick RT DUROID 6035HTC substrate (ɛ_r=3.5 and tanδ=0.0013). The primary dipole element with one tuning inductor in the middle was fabricated on one side while a shorter dipole antenna, which acts as the feeding element, was fabricated on the other side of the substrate (Fig. 1). The total length of the feeding dipole was 70 mm and the length of the EAD passive element was 145 mm.

Electromagnetic simulations were carried out in CST Microwave Studio (CST, Darmstadt, Germany). B₁⁺ and SAR_10g distributions were simulated for a conventional 15 cm straight dipole and for the EAD antenna. The position of the tuning inductor on the conventional dipole was optimized for transmit efficiency.

S₁₁ sensitivity of a single antenna on the load proximity was measured on Vector Network Analyzer (VNA) and a rectangular agarose phantom (ɛ_r=80 and σ=0.4 S/m) at distances from 20 mm to 60 mm.

Eight identical antennas were fabricated and fixed on a plastic cylindrical support with an outer diameter of 26 cm. Each antenna was placed onto the cylinder using a 1 cm thick foam support. A realistic loading phantom (ε_r=50 and σ=0.6 S/m) was centered inside the holder with the antennas.

Measurements were performed on Philips Achieva 7T scanner. In vivo scan were performed on a brain of volunteer with a low tip-angle 3D gradient echo sequence. The sequence parameters were FoV=240x180 mm, slice thickness 3.75mm, FA=10^o, TE/TR=2/4 ms.

Results

Figure 2 shows simulated B₁⁺ profiles of the EAD antenna and conventional dipole. The simulated B₁⁺ field data was normalized to 1W of input power (Fig. 2(c)) and to max SAR_10g (Fig. 2 (d)). In both cases the EAD antenna exhibits slightly higher B₁⁺ efficiency than straight dipole antenna. Figure 3 shows the antenna’s matching sensitivity on the separation distance from the phantom. Figure 4 shows RF shimming with the array when the B₁⁺ fields was focused in three region of interest (ROI) different positions. Figure 5 shows in vivo scans of a human brain with RF shimming with respect to two spatially different regions of interest (Fig. 5 (a) and (b)).

Discussion

Because of the arcs at the end of a passive element of proposed EAD antenna, B₁⁺ longitudinal coverage is slightly longer compared to the straight linear dipole due to a longitudinally more even current distribution on the EAD antenna. A large reduction in SAR compared to equally sized linear dipole while maintaining B₁⁺ efficiency with respect to input power has been observed. The relatively low sensitivity of the S₁₁ on the distance from a phantom ensures low sensitivity of the array on the head size and head position inside an array.

Conclusion

The EAD array element design offers a reduction in SAR compared to equally sized linear dipoles while maintaining B₁⁺ efficiency. RF shimming on a phantom and in vivo also has been demonstrated, indicating the potential of this element for transmit arrays for neuroimaging.

Acknowledgements

References