Purpose: To assess the feasibility to use dynamic hepatocyte-specific contrast-enhanced MRI (DHCE-MRI) as an imaging-based liver function test, and to compare two methods for deconvolutional analysis (DA) in healthy human subjects.
Materials and methods: T(1)-weighted DHCE-MRI with the hepatocyte-specific contrast medium Gd-EOB-DTPA was performed in 20 healthy volunteers. DA was performed using truncated singular value decomposition (TSVD) and Fourier analysis with an appended tail (FA+Tail). Hepatic extraction fraction (HEF) and input relative blood flow (irBF) were calculated for each liver segment. A computer simulation comparing the standard deviation (SD) of TSVD and FA+tail at different levels of signal-to-noise (SNR) ratio was performed. The results obtained were compared using descriptive statistics, the Wilcoxon matched pairs test and the variance ratio test.
Results: Median HEF was 0.201 and 0.205 using TSVD and FA+tail, respectively (P = 0.086). The corresponding results for irBF was 0.240 and 0.239 (P = 0.51). TSVD yielded a smaller SD, although the difference was not significant (P = 0.068 for HEF and P = 0.84 for irBF). The computer simulation showed that TSVD is more stable than FA+tail at most levels of SNR.
Conclusion: DHCE-MRI with Gd-EOB-DTPA enables the calculation of HEF and irBF. We regard these parameters as being markers of hepatic parenchymal function.