Choice of RF coils at 9.4T: SNR and B1+ of transceiver and transmit-only receive-only arrays

Purpose/Introduction: Signal to noise ratio (SNR) and transmit efficiency are the most important considerations in RF coil design. Tight fitting transceiver (TxRx) arrays are often used to optimize the transmit efficiency [1], while large transmit arrays in combination with tight-fitting receive arrays (ToRo) are used for optimum SNR [2]. In this study, a ToRo coil with 8-transmit elements and 18-receive elements was constructed with the aim to approach the transmit efficiency of transceiver arrays while retaining the SNR advantage of receive-only arrays. Subjects and Methods: Experiments were performed on a Siemens 9.4 T whole body scanner. The tight-fitting (20 9 23 cm) 8-channel TxRx array used in this study was presented previously [1]. In the new ToRo coil, the transmit array consisted of eight evenly spaced gapped loops on a circular tube (outer diameter = 26 cm). Each loop was tuned with 12-capacitors and actively detuned with a PIN diode in series. Adjacent elements were inductively decoupled. Both transmit arrays had a length of 10 and 4 cm gap to the shield. The receive part consisted of 18-receive elements arranged in three rows on a helmet (nine in the top row, seven and two in the 2nd and 3rd row, respectively). Adjacent elements in the same row were inductively decoupled and each element of the lower row geometrically decoupled with two elements of the upper row. The full setup is shown in Fig. 1. Results: Coupling between adjacent transmit elements was\-20 dB and active detuning in transmit and receive array was\-30 dB. The measured transmit field closely followed the predicted pattern from simulations whereas the difference in mean B1 + in the central axial slice (B1 + avg) was below 9 compared to the numerical model (Fig. 2a/b). Compared to the TxRx array the B1 + avg of the ToRo configuration was, however, 18 weaker (Fig. 2c). This includes additional losses caused by the receive array and active detuning circuits. Nevertheless, the ToRo coil provided substantially higher SNR than the TxRx configuration (Fig. 3). Discussion/Conclusion: The designed ToRo array provided high B1 + efficiency, but, not to the level of tight fitting TxRx arrays. Since ToRo arrays are loosely coupled to the load, it doesn’t have to be tuned and matched for every experiment. Furthermore, ToRo configuration provides significantly higher SNR and better parallel imaging performance due to smaller coil elements arranged closer to the head. Future studies will include a comparison of B1 +/sqrt(SAR) for the two coil configurations