Quantitative susceptibility mapping (QSM) has been demonstrated as a promising tool for quantifying brain iron concentration in deep gray matter, e.g., normal aging ¹. QSM applications beyond the brain are also under active development, for example, previous studies have been reported the potential applications for iron concentration quantification in human liver². However, both chemical shift and breathing artifacts pose the further challenges for accurate susceptibility measurement in the liver. In this study, chemical shift removal combined with recently developed two-level STAR-QSM methods could reduce the artifacts and allows high quality susceptibility maps of human liver. The purpose of this study was to estimate the susceptibility map in the liver, from which the hepatic iron concentration was measured and further compared with R2-based liver iron concentration (LIC) estimation.

This prospective study was approved by the institutional review board. Informed consent was submitted by all patients. 7 Patients (5 men, 2 women; mean age 44 year) suspected of hepatic iron overload, with an elevated serum ferritin level (1500 mg/L) were scanned on a 1.5 T MR system (Magnetom Aera; Siemens Healthcare, Erlangen, Germany) using a 3D multi-echo GRE sequence with the following parameters: TR=10 ms; TE1/spacing/TE6=1.44/1.36/8.24 ms; bandwidth=1040 Hz/pixel; flip angle=6°; FOV=420 x 315 mm2; matrix size =416 X 384; slice thickness = 5 mm. 2D data sets for Ferriscan LIC measurment were acquired using a 2D spin echo sequence with the following parameters: TR =1000 ms; TE1/spacing/TE5=6/3/18 ms; bandwidth = 501 Hz/pixel; FOV = 400 x 300 mm2; matrix size=256 x 192; slice thickness =5 mm. The acquired 2D SE MR data were uploaded to Resonance Health (Claremont, Australia) for further Ferriscan LIC analysis.

The phase was unwrapped using Laplacian-based unwrapping and background phase was removed using V-SHARP3. The QSM images were reconstructed using a two-level streaking artifact reduction for QSM algorithm4. Liver susceptibility values were calculated relative to that of the subcutaneous adipose tissue as which doesn’t accumulate excess iron. The computed QSM images were compared with R2* maps. Approximate slices were selected according to the corresponding cross-section on the LIC report. The ROIs were manually drawn on QSM images according corresponding LIC maps. Linear regression analysis was performed to investigate the correlation between liver susceptibility and the Ferriscan-based estimate of LIC.

Fig.1 showed the magnitude, R2* and susceptibility maps for two subjects with different levels of iron concentration. As the hepatic iron deposition increased, the susceptibility in the liver increased more paramagnetic (positive susceptibility). The iron deposits were in good agreement with R2* as the iron has longer R2* (1/T2*). As expected, liver susceptibility value increased with increasing iron overload, which was confirmed by Ferriscan LIC (Fig.2). The Ferriscan LIC and mean susceptibility values were 0.9, 5.9, 10.2 (mg/g dry) and 0.06, 0.1, 0.29 (ppm) for the three subjects, respectively. Significant positive correlation was observed between QSM and Ferriscan LIC based iron measurement (Fig.3) with the coefficient of determination R2 = 0.8.

In this study, we have demonstrated that high quality susceptibility maps can be used to measure the different iron levels of liver. Liver susceptibility demonstrated a good correlation with liver iron deposition measured using Ferriscan LIC (R2 = 0.8). High correlation is due to both the chemical shift removal and the promising two-level QSM reconstruction algorithm.

Hepatic iron overload causes oxidative hepatocellular injury and progressive fibrosis, then developing into cirrhosis and hepatocellular carcinoma 5,6. Excessive hepatic iron even might lead to cardiac complications and early death7. Accurate assessment is essential for the treatment with iron chelators to guarantee suitable chelator dose, avoiding both toxicity from iron overload and side-effects from an excessive chelator dose8. Empirically derived calibrations reflect the uncertainty of R2*. As for Ferriscan LIC, the high price and long examination time restrict its wide use. The advantage for QSM is that susceptibility is a fundamental property of materials, with a single breath hold scanning time and appropriate price.

Note that in this study, ROIs in single slice was evaluated both in QSM and Ferriscan LIC instead of 3D large volume.

In conclusion, high quality QSM maps may allow more reliable estimates of iron deposition in the liver.