PHASE UNWRAPPING IN THE 3-POINT DIXON METHOD FOR FAT-SUPPRESSION MR-IMAGING

PURPOSE: To present an algorithm for performing phase unwrapping for fat and water signal separation in the three-point Dixon (3PD) method of magnetic resonance (MR) imaging. MATERIALS AND METHODS: The algorithm is based on a two-dimensional region-growing approach, which tracks phase evolution in 3PD images and unwraps the phase when a 2 pi jump is detected. The unwrapped phase data are consecutively used to calculate fat and water signal components on a per-pixel basis. The method was tested in eight volunteers. RESULTS: The algorithm creates four output images: standard spin-echo, fat suppression, water suppression, and phase map images. This method corrects for the effects of magnetic field inhomogeneities and provides spatially uniform fat signal suppression superior to that achievable with the standard method. CONCLUSION: This method of 3PD data reconstruction appears to be promising for routine application of fat suppression MR imaging in a clinical setting.