Comparison of Quantitative Susceptibility Mapping algorithms based on numerical and in vivo 3T data

This study compares the currently publically available algorithms for quantitative susceptibility mapping, including different phase unwrapping, background field removal and dipole inversion methods. Numerical and human in vivo brain MRI data are used for a qualitative and quantitative assessment of the various methods. In 3T in vivo MRI data, phase unwrapping with combined spatial and temporal fitting and background field removal using V-SHARP results in the least artifacts. MEDI and iLSQR are currently the most accurate dipole inversion algorithms, with a significantly shorter processing time for the iLSQR method.