Laterally Oscillating Trajectory for Undersampling Slices (LOTUS)

Motivation: Simultaneous multislice (SMS) MRI greatly improves scan efficiency, but there has been little investigation of optimal non-Cartesian trajectories for SMS. Goal(s): To introduce Laterally Oscillating Trajectory for Undersampling Slices (LOTUS) — a new single-shot spiral-like trajectory for SMS that efficiently covers 3D k-space pseudo-randomly, which may be advantageous for compressed sensing (CS). Approach: Trajectory performance was assessed using both simulation and in vivo MRI in human participants. Results: Simulations show reduced errors in CS reconstruction and lower g-factor when using proposed trajectory compared to alternatives. In-vivo results show significantly improved noise variance with high slice accelerations. Impact: We introduce a k-space trajectory that enables higher slice accelerations compared to traditional approaches, especially compared to Blipped-CAIPI EPI. This may be useful to reduce scan times for advanced diffusion MRI acquisitions or improve temporal resolution of functional MRI.