AIM: To evaluate the relationship between histopathology results and magnetic resonance imaging (MRI) on in and opposed-phase sequences grading of fat deposition within human liver. MATERIALS AND METHODS: In and opposed-phase T1-weighted gradient-echo sequences (double echo time 2.3 ms and 4.6 ms) were performed in 25 patients, using a 1.5-T clinical MR imaging system. Fat/water ratio on in- and opposed-phase images of the liver was compared with pathologically defined degree of steatosis. The signal intensity in the images was acquired with operator-defined regions of interest at the same location in both fat and water images and the ratio was calculated by dividing signal intensity of liver in opposed phased sequence on signal intensity of liver in phased sequence. Fat/water ratio and the degree of steatosis were compared using linear regression. The sensitivity and specificity of opposed-phase for diagnosing steatosis were defined by ROC analysis. Furthermore, a correlation between visual signal intensity variation and the degree of steatosis was assessed using Pearson correlation coefficient. RESULTS: Histology demonstrated fatty liver infiltrations in 81% of specimens. The percentage of fatty hepatocytes was 28 +/- 30%. Fat/water ratio was significantly correlated with the pathologic grading of steatosis (r = 0.816, P < 0.001). The opposed phase MR imaging sensibility and specificity for the diagnosis of hepatic steatosis were respectively 80% and 71%. We obtained a statistically significant correlation between visual SIV and fatty liver grading (P = 0.017). CONCLUSION: We demonstrated a significant correlation between fat/water ratio and histological findings for the detection and grading of fatty liver.
AIM: To evaluate the relationship between histopathology results and magnetic resonance imaging (MRI) on in and opposed-phase sequences grading of fat deposition within human liver. MATERIALS AND METHODS: In and opposed-phase T1-weighted gradient-echo sequences (double echo time 2.3 ms and 4.6 ms) were performed in 25 patients, using a 1.5-T clinical MR imaging system. Fat/water ratio on in- and opposed-phase images of the liver was compared with pathologically defined degree of steatosis. The signal intensity in the images was acquired with operator-defined regions of interest at the same location in both fat and water images and the ratio was calculated by dividing signal intensity of liver in opposed phased sequence on signal intensity of liver in phased sequence. Fat/water ratio and the degree of steatosis were compared using linear regression. The sensitivity and specificity of opposed-phase for diagnosing steatosis were defined by ROC analysis. Furthermore, a correlation between visual signal intensity variation and the degree of steatosis was assessed using Pearson correlation coefficient. RESULTS: Histology demonstrated fatty liver infiltrations in 81% of specimens. The percentage of fatty hepatocytes was 28 +/- 30%. Fat/water ratio was significantly correlated with the pathologic grading of steatosis (r = 0.816, P < 0.001). The opposed phase MR imaging sensibility and specificity for the diagnosis of hepatic steatosis were respectively 80% and 71%. We obtained a statistically significant correlation between visual SIV and fatty liver grading (P = 0.017). CONCLUSION: We demonstrated a significant correlation between fat/water ratio and histological findings for the detection and grading of fatty liver.
Authors: Larry Cai; Benjamin M Yeh; Antonio C Westphalen; John P Roberts; Zhen J Wang Journal: Diagn Interv Radiol Date: 2016 May-Jun Impact factor: 2.630