Phase Correction Algorithms for Water/Fat-Separation in MRI

The quality of water/fat-separation in magnetic resonance imaging (MRI) strongly depends on the result of the phase correction procedure. A completely satisfactory solution for noisy complex data is still not available. The intention of the current work was to provide a two-step phase correction region-growing algorithm for water/fat-separation.
The underlying phase correction process is implemented as a two-step process correcting the linear component of the phase errors with a modified Ahn-Cho algorithm first and then corrects the residual phase errors with a region-growing algorithm.

The algorithm is implemented in C and tested on phantoms as well as on in-vivo data. Representative results of the water/fat-separation are shown and compared to the Two-Point Dixon method.
Advantages and drawbacks of the region-growing algorithm are evaluated, the fundamental limitations and possible improvements are reported and discussed.