Clinical evaluation of nonalcoholic fatty liver disease with in-phase and out-of-phase subtraction MR imaging and T2 corrected proton 1H-MR spectroscopy
SHI Yu, GUO Qi-yong, XIA Fei, ZHANG Lan, BIAN Sheng-xin
Shengjing Hospital, China Medical University, Shengyang 110004, China
Abstract:Objective: To compare the T1 weighted in-phase and out-of-phase magnetic resonance imaging(IP-OP) with subtraction techniques and T2 corrected proton 1H-MR spectroscopy(1H-MRS) in the assessment of liver fat content(LFC) in patients with nonalcoholic fatty liver disease. Methods: IP-OP MR imaging and 1H-MRS were performed in 42 patients using a 3T MR unit. Subtraction images were made from pairs of IP and OP images. T2 relaxation measurements were performed in five patients by using a multi echo PRESS sequence. LFC measured by IP-OP and 1H-MRS were compared with histological results using spearman analysis and Mann-Whitney U test. Results: LFC measured by 1H-MRS were corrected by T2 relaxation(T2water=33.1 ms, T2fat=64.2 ms). Sensitivity and specificity of IP-OP substraction and 1H-MRS to detect steatosis was 89.2%, 100% and 91.2%, 100%, respectively. IP-OP MR imaging and 1H-MRS measurements of LFC had strong correlation with histopathologic steatosis assessment(r=0.80, 0.92; P<0.01). IP-OP and 1H-MRS showed differences across steatosis grades: none versus mild, mild versus moderate and moderate versus severe grade(expect that IP-OP cannot differentiate moderate from severe grade, Z≤-2.54, P<0.01). The LFC measured by IP-OP and 1H-MRS were significantly lower than pathology(expect for non fatty liver group, Z≤-2.21, P<0.05). Conclusion: IP-OP is a quick and effective way to detect steatosis for clinic. 1H-MRS has better diagnostic accuracy in depicting hepatic steatosis than IP-OP.