Phantom Experiments for Magnetic Susceptibility

The phantom experiments utilized a gadolinium phantom consisting of six 50 mL centrifuge tubes (Corning, Corning City, New York) in a water bath. Each tube contained a solution with different concentrations of gadobutrol (Bayer, Leverkusen, Germany). The concentrations were chosen to span the range of physiological magnetic susceptibilities arising from excess iron [6] (link) and include higher values to assess the effect of streaking artifacts at higher susceptibility shifts. The concentrations were 0.0, 0.6, 1.5, 3.1, 6.1, 9.2 mM/dm³ giving theoretical susceptibilities of 0.0, 0.2, 0.5, 1.0, 2.0, 3.0 ppm, respectively, when calculated with the Curie law [17] (link). The shim-box and reconstruction ROI were set to the volume of the water bath. The mean and standard deviation of the voxels within an ROI defined for each tube in a central slice (to avoid the large susceptibility shift relative to water, of the acrylic tube stand which held the tubes at the top and bottom) of the resultant susceptibility map were calculated, and fit with a linear model of measured against predicted susceptibility. λ values of 50, 100, 250, and 500 were used for the experiment and L-curve analysis was performed using the zero curvature approach [18] (link) across a range of λ values from 0.1 to 1 × 10⁷.