Introduction:

Tight-fit human head ultra-high field (UHF, 7T and above) transceiver (TxRx) surface loop phased arrays (1,2) improve transmit (Tx) efficiency in comparison to Tx-only arrays, which are usually larger to fit multi-channel receive (Rx)-only arrays inside (3). A drawback of the TxRx-design is that the number of array elements is restricted by the number of available radiofrequency (RF) Tx-channels (commonly 8 or 16). This channel count is not sufficient for an optimal SNR and parallel Rx-performance. In this work we developed a method of increasing the number of Rx-elements in a TxRx-array without the necessity of moving the Tx-elements further away from the subject, which compromises the Tx-performance. We designed and constructed a prototype of a 16-channel human head phased array, which consists of 8 TxRx-surface “horizontal” loops circumscribing a head, and 8 Rx-only “vertical” loops positioned along the central axis (parallel to B₀) of each TxRx-loop perpendicularly to its surface.

Methods:

The array consists of 8 overlapped horizontal (10.5 cm in width; 10 cm in length) TxRx-loops, and 8 vertical (3.3 cm x 10 cm) Rx-only loops (Fig.1) and measures 20 cm in width (left-right) and 23 cm in height (anterior-posterior). Overlapping of adjacent horizontal loops provides excellent decoupling below -30 dB (Fig.2) (4). All vertical loops are decoupled by preamplifier decoupling during reception and actively detuned during transmission. The array is shielded with a cylindrical shield at 4-cm distance to the coil elements. Electromagnetic simulations of the transmit B₁⁺ field and the local specific absorption rate (SAR) were performed using CST Studio Suite 2015 (CST, Darmstadt, Germany) and the time-domain solver based on the finite-integration technique (FIT). We also included vertical Rx-only loops into simulations since the presence of actively detuned Rx-only elements may substantially alter the maximal local SAR (5). Three voxel models were used, i.e. a head/shoulder (HS) phantom (3), which was constructed to match tissue properties (ε = 58.6, σ =0.64 S/m at 400 MHz) (3), and two virtual family multi-tissue models, “Duke” and “Ella”. Experimental B₁⁺ maps were obtained using the AFI sequence (6). All data were acquired on a Siemens Magnetom 9.4T human imaging system.

Results and Discussion:

First we performed a safety evaluation of the array including EM modeling (B₁⁺, SAR) and phantom imaging. Simulations showed that the maximum local SAR was increased by less than 3%, and B₁⁺ efficiency decreased by less than 1% due to introduction of the vertical loops. Then we conducted in-vivo experiments (Fig.3). Averaged over the 20-mm central slab B₁⁺ measures 12.4 ± 2.7 μT/√kW in case of the 16-channel design versus 12.5 ± 2.8 μT/√kW in case of the 8-channel design. Introduction of the vertical Rx-only loops improved the in-vivo SNR near the coil center by ~28% and at the periphery by ~30% to 40%. Finally we compared parallel Rx-performance of the 16-channel array with that obtained using 8 horizontal loops only (Table 1). As an example Fig.4 demonstrates transversal G-factor maps obtained near the array center. At high acceleration (AF=5) the 16-channel mode improves an average G-factor up to 20% for an acceleration in the AP-direction. An idea of combining vertical and horizontal loops is similar to a recently proposed method of combining dipole antennas with surface loops to provide for better SNR (7). According to an ultimate intrinsic SNR approach, a combination of the curl-free and divergence-free currents on a cylindrical surface is required to produce an optimal SNR (8,9). In our case the curl-free current component is produced by vertical loops, which are easier to construct in terms of matching, tuning, decoupling, and active detuning, than dipole antennas.

Conclusion:

As a proof of concept we developed a method of increasing the number of Rx-elements in a human head transceiver array without the necessity of moving the transmit loops away from the subject head and thus compromising the Tx-efficiency. We constructed a prototype of a 16-channel 9.4T tight-fit head phased array, which consists of 8 horizontal TxRx-surface loops circumscribing a head, and 8 Rx-only vertical loops placed at the center of each TxRx-loop. We demonstrated both experimentally and numerically that addition of the vertical loops substantially improves the Rx-performance without compromising the Tx-performance and the maximum local SAR. To further improve SNR and the longitudinal coverage along the array axis, increasing the overall number of array elements, (i.e. double-row 16-TxRx-element and 32-Rx-element array) is required.

Acknowledgements

No acknowledgement found.

References

References: 1) Gilbert KM et al, MRM 2010;67:1487. 2) Avdievich NI et al, Proc.ISMRM 22,2014,622. 3) Shajan G et al, MRM 2014;71:870. 4) Avdievich NI et al, Proc.ISMRM 24,2016,169. 5) Krishnamurthy N et al, JMRI 2014;39:475. 6) Yarnykh VL. MRM 2007;57:192. 7) Chen G et al, Proc.ISMRM 23,2015,3133. 8) Lattanzi R et al, MRM 2012;68:286. 9) Pfrommer A et al, Proc.ISMRM 24,2016,175.