Motivation: Fast imaging sequences exploit higher gradient amplitudes and slew rates and efficient non-Cartesian k-space trajectories. With these improvements comes greater concern for PNS, acoustic noise, and system damage. Goal(s): To design gradient waveforms which efficiently follow an arbitrary k-space trajectory while controlling time varying properties such as PNS, and frequency content. Approach: We construct and solve an optimization problem for k-space speed across trajectory arclength. This determines the trajectory's time parameterization, and therefore the gradient waveform properties. Results: We demonstrate 10% faster PNS-limited readout than conventional methods, up to 9dB decrease in acoustic noise, and up to 19x reduction in gradient coil vibrations. Impact: We develop a novel customizable gradient waveform design method (OPTIKS) capable of optimizing time-dependent properties while adhering to any specified trajectory. We apply OPTIKS to design fast-PNS limited readouts, make imaging quieter, and reduce large prolonged gradient coil vibrations.
McCready et al. (Tue,) studied this question.