Item

Abstract

Purpose/Introduction: Ultra-high field (UHF) (C7 T) tight fit transceiver head phased arrays improve Tx-efficiency (1, 2) in comparison to larger Tx-only arrays (3). However, increasing the number of elements above 16 is difficult due to decoupling problems and the limited number of Tx-channels. The number of Rx-elements needs to be higher to improve SNR and parallel Rx-perfomance (3). In this work, we developed and constructed a novel splittable transceiver array at 9.4 T (400 MHz). The array doubles the number of Rxelements while keeping both Tx- and Rx-elements at the same distance to a subject. Both Tx and Rx-performance are optimized using this technology. Subjects and Methods: The array (Fig. 1) consists of 4 Tx-loops (8 9 9 cm). Each loop can be split into two smaller Rx-loops (Fig. 1b). Splitting is provided by 4 PIN diode switches, S1, S2. During transmission four larger loops are produced by shortening S1 and opening S2 switches. Open S1 and shortened S2 switches produce two Rx-loops during reception. All switches are connected in series and driven by 100 mA current. The array can also be used in 4-channel mode without splitting. In both modes a 4-way splitter with 1008 phase shift between the loops was used during transmission. B1 + maps were obtained using the AFI sequence (4) and a rectangular phantom (26 9 20 9 12 cm) matching tissue properties (3). SNR and g-factor maps were obtained using non-accelerated and GRAPPA accelarated GREs (3). Data were acquired on the Siemens Magnetom Discussion/Conclusion: Both SNR and parallel reception were substantially improved by doubling the number of Rx-elements. Arrangement of elements can be adapted to future head or body applications. The direction of splitting can be chosen depending on the most beneficial direction of acceleration. Tx-elements splittable into a larger number of Rx-elements (4, 8) would allow acceleration in both directions. As a proof of concept we developed a novel splittable transceiver phased array. Both Tx- and Rx-performance are optimized using this method.