This project adapts the SMURF measurement-model approach to 2D phase-contrast MRI. The goal is to infer a geometry representation and a velocity field on 2D slices while accounting for phase artifacts and noise.
Work includes defining the measurement likelihood for magnitude and phase-derived velocities in 2D, implementing inference with subsampling, and producing outputs that can be used for downstream analysis. Extensions include multi-slice aggregation and consistency checks on reconstructed flow fields.