A Fully Automatic Pipeline to Optimize Radio-Frequency Coil Design Using the Ultimate Intrinsic Signal-to-Noise Ratio as the Benchmark

Motivation: Despite the crucial importance of radio-frequency coils, the practical coil design process has remained a largely empirical one. Goal(s): To introduce a fully automatic procedure to optimize coil design using the ultimate intrinsic signal-to-noise ratio (UISNR) as the reference. Approach: We developed an intuitive user interface to parameterize and mesh coils, and implemented automatic tune-match-decoupling. We used MARIE to calculate coil SNR and the UISNR inside a head model. We optimized the size and position of loop coils with the goal of maximizing SNR/UISNR at target locations. Results: The optimization converged to coil configurations resembling the ideal current patterns. Impact: Optimizing the design of receive coil arrays can improve image quality and parallel imaging performance. This work demonstrates the first fully automatic approach to design receive coils that can approach the ultimate intrinsic signal-to-noise ratio.