Quantitative susceptibility mapping (QSM) has become a robust technique to measure voxel-wise tissue magnetic susceptibility from phase images. QSM can provide in vivo quantitative susceptibility value¹². Gadolinium-based contrast agents (GBCAs) have been widely used for contrast material-enhanced magnetic resonance (MR) imaging. However, gadolinium accumulation in the dentate nucleus (DN) has gained attention due to recent studies^3-6. These recent studies have focused T1 shortening effect of gadolinium in the DN on T1-weighted images, and susceptibility changes have not been evaluated due to a paramagnetic effect of GadoliniumSubjects We retrospectively enrolled 49 patients who had underwent contrast enhanced-MR imaging and 49 healthy volunteers to establish normal database for QSM. MR imaging All MR scans were performed at3T MR scanner (Magnetom Skyra, or Trio, Siemens, Erlangen, Germany). Three-dimensional (3D) axial gradient-echo sequence (repetition time (TR) / echo time (TE) /ΔTE, 55 ms / 3.6-45.0 ms / 5.91 ms ; resolution; 0.9×0.9×2.0 mm) and 3D sagittal T1-weighted image (VIBE: TR/TE, 6.0/2.29; flip angle, 15°; resolution, 0.9×0.9×0.9, or MPRAGE: TR/TE/TI, 1900/2.58/900; flip angle, 9°; resolution, 0.9×0.9×0.9) were used. 3D sagittal T1-weighted images were reconstructed to axial plane parallel to the axial images manually. Post-processing QSM calculation was conducted from the magnitude and phase images of the gradient echo images by STI Suite version 2.10 (http://people.duke.edu/~cl160/). At first, we conducted phase unwrapping and background phase removal using the sophisticated harmonic artifact reduction for phase data with a variable radius of the spherical kernel at the brain boundary (V-SHARP method 78. Then, QSM was reconstructed from the local tissue phase images by solving an inverse problem using the algorithm for sparse linear equations and sparse least squares (iLSQR method)7 . Data analysis A region of interest (ROI) study was performed. The mean of the susceptibility values of the DN were measured and T1 ratio was defined as follows: mean values of the DN divided by those of the cerebellar white matter. Statistics We performed the unpaired t test with the Welch correlation to assess the difference of susceptibility values and T1 ratio between GBCA group (patients) and non-GBCA group (healthy volunteers). We also assessed the regression analysis between susceptibility values or T1 ratio and the number of the GBCA administration, and the regression analysis between susceptibility values and the T1 ratio. All the statistical analysis were conducted using MedCalc version 13.3 (MedCalc Software bvba, Ostend, Belgium). P values of less than .05 indicated a statistically significant difference.

The subjects’ characteristics are shown in Table 1. There were no patients with renal dysfunction. The age distribution between the GBCA group and non-GBCA group was not significantly different (P=0.317). Typical images of QSM and T1-weighted images are shown in Figure 1.

Susceptibility value and T1 ratio between GBCA and non-GBCA group The susceptibility values at DN in GBCA group was significantly higher than those of non-GBCA group (GBCA group, 0.187ppm±0.031, non-GBCA group, 0.081ppm±0.026, P<0.0001) (Fig. 2). T1 ratio at DN in GBCA group was also significantly higher than that in non-GBCA group (GBCA group, 1.053±0.071, non-GBCA group, 0.989±0.023, P<0.0001) (Fig. 3)

QSM could detect the susceptibility change due to paramagnetic substance. Gadolinium is the very strong paramagnetic substance. In current study, we focused on the gadolinium deposition in the DN. Gadolinium usually never exist in the human body, but Kanda et al demonstrated that Gadolinium deposition in the DN due to serial GBCA administration. This phenomenon were reported to be observed as high intensity of T1 weighted images in the DN. Susceptibility change of GBCA group against non-GBCA group reflect the gadolinium deposition due to serial GBCA administration. the result of our study also reconfirmed the accumulation of Gadolinium in the DN, specifically the higher susceptibility values in the DN on the GBCA group. QSM can also detect the susceptibility changes in dentate nucleus due to gadolinium deposition from serial enhanced MR scans.