Notes on quantitative susceptibility mapping reconstruction accuracy under challenging conditions: Phantom measurements and simulations

Abstract

Magnetic susceptibility can be assessed by quantitative susceptibility mapping (QSM), based on measured magnetic resonance imaging (MRI) phase data. The QSM reconstruction process is, however, mathematically challenging and still not fully robust. A signal-generating holmium [Ho(III)] aqueous solution with air-equivalent magnetic susceptibility was prepared, and used as a surrounding medium in a water phantom with tubes filled with a solution of gadolinium contrast agent at various concentrations. Extended analyses under controlled conditions were accomplished by simulations of the phantom construction. Without surrounding holmium solution, a gadolinium tube positioned centrally, parallel with B0, showed a susceptibility difference that agreed well with theoretical values, whereas a peripheral parallel tube position showed larger deviation. Orientation perpendicular to B0 resulted in less variation between the internal tube positions. Air-equivalent magnetic susceptibility corresponded to 16.5 mM Ho(III) solution. With surrounding holmium solution, several post-processing steps became challenging. Simulations indicated higher degree of underestimation when the theoretical susceptibility difference increased. Details in the mathematical implementation, for example, background field removal can strongly influence the result. Simulated results were, in part, unexpected, and provided awareness of limitations in the reconstruction technique, mainly related to conditions with large susceptibility differences between compartments.