Abstract

It is well known that reconstruction algorithms in quantitative susceptibility mapping often contain streaking artifacts. These are nondesirable objects that contaminate the image, and the possibility of removing or at least reducing them has a great practical interest. In [J. K. Choi, H. S. Park, S. Wang, Y. Wang, and J. K. Seo, SIAM J. Imaging Sci., 7 (2014), pp. 1669-1689], the cause of the artifacts is identified as propagation of singularities for a wave-type operator. In this work, we analyze such singularities using microlocal techniques and propose some strategies to reduce the artifacts.