Objective To investigate the diagnostic value of proton magnetic resonance spectroscopy （1 H-MRS） on the quantification of hepatic steatosis at 3.0 T MR united.Methods Twenty-two patients who were candidated for liver surgery (living liver transplantation donor candidates, lobectomy or segmental resection for focal liver diseases, etc.) were enrolled in this study. 1 H-MRS was conducted with point resolved selective spectroscopy （PRESS） sequence, using SAGE software packages. The values of water peak (PW), lipid peak (PL) were measured, the area under water peak (AW) and lipid peak (AL), and then the relative lipid content of liver cells (RLC1, RLC2) was calculated. All subjects underwent surgical resection of liver shortly after MR scanning, fresh frozen specimens were obtained for Sudan Ⅲ staining, and staging was conducted.Results In all 22 patients, 7 patients without fatty liver, 11 patients with mild fatty liver, 4 patients with moderate or severe fatty liver. Intercomparison between different histopathological grades showed following findings: the values of PL, AL, RLC1 and RLC2 had statistical significance (Plt;0.05). The values of PL， AL， RLC1 and RLC2 were positively correlated with the proportion of fatty degenerative cells (PFDC), Plt;0.05, and the highest correlation factor was 0.771 (RLC1 to the PFDC).Conclusion 1 H-MRS can accurately quantify liver fat content and reflect the severity of hepatic steatosis. It has the potential to replace invasive liver biopsy.
ObjectiveTo investigate the diagnostic value of spectral saturation inversion recovery, gradient-echo chemical shift MRI， and proton magnetic resonance spectroscopy in quantifying hepatic fat content. MethodsConventional T1-weighted and T2-weighted scanning (without fat saturation and with fat saturation)， gradient-echo T1W in-phase （IP） and opposedphase （OP） images and 1H-MRS were performed in 31 healthy volunteers and 22 patients who were candidates for liver surgery. Signal intensities of T1WI amp; T1WIFS (SInonfat1, SIfat1), T2WI amp; T2WI-FS (SInonfat2, SIfat2), and IP amp; OP (SIin, SIout) were measured respectively, the relative signal intensity one (RSI1), relative signal intensity two (RSI2)， and fat index (FI) were calculated. Peak values and the area under peak of 1H-MRS were measured, and the relative lipid content of liver cells (RLC ) were calculated. Twenty-two patients accepted liver resection and histological examination after MRI scanning, the proportion of fatty degenerative cells were calculated by image analysis software. Results①Hepatic steatosis group showed higher average values of RSI1, FI， and RLC to non-hepatic steatosis group (Plt;0.05), while there was no significant difference in RSI2 between two groups (Pgt;0.05). ②There was a statistical significant difference in RLC among different histopathological grades of hepatic steatosis, and RLC increased in parallel with histopathological grade (Plt;0.05).There was no significant difference in RSI2, RSI1, and FI among different histopathological grades, although the latter two had a tendency of increasing concomitant with histopathological grade (Pgt;0.05). ③The values of FI and RLC were positively correlated with the PFDC （r=0468, P=0.027； r＝0771， Plt;0.000 1）, while they were not in RSI1 and RSI2 (r=0.411, P=0.057； r=0.191, P=0.392). ConclusionsSPIR, Gradient-echo chemical shift MRI and 1H-MRS can help to differentiate patients with hepatic steatosis from normal persons, the latter also can help to classify hepatic steatosis. In quantifying hepatic fat content, 1H-MRS is superior to gradient-echo chemical shift MRI, while SPIR’s role is limited.