MRI Pulse Sequence Discovery with Joint Optimization of B-spline parameterized Trajectory and Sequence Parameters

Motivation: Existing gradient-based methods for sequence discovery may yield impractical kspace trajectories not realizable on MRI hardware. Goal(s): Optimization of non-Cartesian k-space trajectories, flip angles, and sampling times using B-splines parameterization and backpropagation. Approach: We parameterized radial k-space trajectories with B-splines for smooth manipulation. Joint optimization of B-spline coefficients, flip angles, and sampling times was achieved through minimizing reconstruction loss against referece reconstrcution and hardware constraints. Results: This approach yields unconventional sequences with smooth non-Cartesian k-space trajectories, that closely approximates the fully sampled FLASH reference. Impact: This method enables design of optimized MRI sequences tuned to specific imaging objectives, opening possibilities for application-specific sampling patterns and optimized sequences across diverse imaging tasks.