Objective: To evaluate the value of dual-energy CT (DECT) with use of an iron-specific, three-material decomposition algorithm for the quantification of liver iron content (LIC).
Methods: Thirty-one phantoms containing liver tissue, fat and iron were scanned with dual-source CT using single-energy at 120 kV (SECT) and DECT at 80 kV and 140 kV. Virtual iron concentration (VIC) images derived from an iron-specific, three-material decomposition algorithm and measurements of fat-free and fat-containing phantoms were compared with the LIC and healthy liver tissue.
Results: In the absence of fat significant linear correlations were found between LIC and HU from SECT and VIC (r = 0.984-0.997, p < 0.001) with a detection limit of 145.4 μmol/g LIC for SECT, whereas VIC accurately quantified the lowest LIC of 20 μmol/g dry liver. In the presence of fat, no significant correlation was observed between LIC and SECT, whereas significant correlations were found for VIC. Compared with fat-free phantoms, significant underestimation of LIC was seen for SECT with increasing amounts of fat (all, p ≤ 0.01). On the other hand, similar HU were seen for VIC of fat-containing compared with fat-free phantoms (p > 0.632).
Conclusions: Virtual iron concentration images generated from DECT provide added value for the quantification of LIC by disregarding the confounding effect of the natural variation of healthy liver attenuation and of co-existing liver fat.