Design and Implementation of Two 16-Element Antisymmetric Transceiver Coil Arrays for Parallel Transmission Human Cardiac MRI at 7 T

Abstract

Two 16-element radio frequency (RF) transceiver antisymmetric coil arrays with two different implementation designs aiming for improved characteristics for static phase $B_{1}^{+}$ shimming and parallel receive were developed for human cardiac MRI at 7 T. Both cardiac arrays consist of a weakly curved lightweight housing fit to an average human thorax shape as an anterior array in combination with a flat posterior array. The first array (Design 1) comprised eight identical antisymmetric loops for both anterior and posterior sections. Based on initial testing and to improve parallel imaging capabilities and $B_{1}^{+}$ field homogeneity, the second array (Design 2) was composed of 12-loops for the anterior section and four antisymmetric loops of larger size for the posterior section. Electromagnetic-field (EMF) simulations were carried out for both antisymmetric arrays loaded with an elliptical-shaped thorax phantom and two human voxel models (Duke and Ella). Static phase $B_{1}^{+}$ shimming has been carried out within Duke and Ella models using two optimization cost functions aiming to maximize the transmit efficiency and weighted combination of both $B_{1}^{+}$ field homogeneity and efficiency. The hardware and imaging performance of the two developed antisymmetric cardiac arrays was validated through EMF simulations, benchtop measurements, and MR measurements in a thorax phantom. Both antisymmetric arrays were compared to two commercial 16-element transceiver arrays (a 1Tx/16Rx in single transmit mode and an 8Tx/16Rx in parallel transmit mode). MRI measurements were performed with Design 2 using a 70-kg fresh pig cadaver (10–15 min postmortem). Parallel imaging with an acceleration factor up to $R = 6$ was possible using Design 2 while maintaining a mean g-factor of 1.47 within the pig heart. $T_{2}^{\ast }$ -weighted images of the pig heart were acquired using up to $R = 5$ with a spatial resolution of $0.35 \times 0.35 \times 4$ mm3.