Study On The Evaluation Of Major Features Of Liver Imaging Reporting And Data System By Hepatocyte-specific Contrast Agent Gd-EOB-DTPA

Part 1 Evaluation of the major features of Liver Imaging Reporting and Data System based on subtraction techniqueBackground and Objective:The major features of Liver Imaging Reporting and Data System（LI-RADS）include:nonrim arterial phase hyperenhancement（Nonrim APHE）,nonperipheral washout,enhancing capsule,size and threshold growth.The background signals of liver lesions and surrounding liver parenchyma will affect the evaluation of the major features.For example,benign liver lesions with hyperintensity of T₁WI may show hyperintensity in arterial phase and be misjudged as Nonrim APHE.And the hepatocellular carcinoma（HCC）with hyperintensity on T₁WI,even with nonperipheral washout,may present as isointensity or hyperintensity rather than hypointensity during the portal venous phase.MRI subtraction technique can eliminate the influence of background signals of liver and lesions,and help to correctly evaluate the signal changes of lesions before and after enhancement.In this thesis,the value of subtraction technique in evaluating the major features of LI-RADS v2018 on Gd-EOB-DTPA dynamic contrast-enhanced MRI and its influence on the classification of liver lesions were explored.Materials and methods:The Gd-EOB-DTPA enhanced MRI of 117 histologically verified HCCs in 87high-risk patients was analyzed retrospectively.For the conventional and subtraction images,two radiologists independently evaluated the major features:Nonrim APHE,nonperipheral washout,enhancing capsule,and graded the lesions with LI-RADS v2018 as the standard.According to the lesion size（<20 mm,≥20 mm）and T₁-weighted signal intensity（hypointensity,isointensity or hyperintensity）,the patients were divided into different subgroups.The detection rate of major features in conventional images and subtraction images were counted.The paired sample chi-square test（Mc Nemar test）was used to compare whether the difference between the detection rate of the major features of the conventional and subtraction images and the diagnosis rate based on LR-5 as the diagnostic standard was statistically significant.Results:For all HCC（89.74%vs 75.21%,P=0.0002）,hypointensity HCC（90.00%vs70.00%;P=0.0001）and HCC≥20 mm（92.31%vs 76.92%;P=0.0005）,the detection rate of arterial phase hyperenhancement of subtraction images from arterial phase was significantly higher than that of conventional arterial images.For HCC<20 mm,the conventional arterial phase image combined with subtraction images from arterial phase can significantly improve the detection rate of arterial phase hyperenhancement than conventional arterial phase image（89.74%vs 71.79%;P=0.0156）.Only the subtraction image from arterial phase is used（84.62%vs 71.79%,P=0.1797）,the detection rate of arterial phase hyperenhancement did not increase significantly.For isointensity or hyperintensity HCC,the combined subtraction image from portal venous phase can significantly improve the detection rate of nonperipheral washout（75.68%vs 54.05%;P=0.0078）.No matter in all HCCs or subgroups,the detection rate of enhancing capsule of subtraction images from portal venous phase was higher than that of conventional portal venous phase（P<0.05）.The LR-5 diagnosis rate of HCC was significantly improved after the conventional image combined with subtraction image from arterial phase（71.79%vs 56.41%,P=0.0001）or subtraction image from portal venous phase.（66.67%vs 56.41%P=0.0075）.Conclusion:Subtraction imaging on Gd-EOB-DTPA dynamic contrast-enhanced MRI can upgrade LI-RADS categories for the diagnosis of HCC because of its superior ability in depicting Nonrim APHE,nonperipheral washout and enhancing capsule.Prat 2 Study on the consistency of liver lesion size measured in different phases by gadolinium disodium dynamic contrast-enhanced MRIBackground and Objective:The size of liver lesions is very important for the diagnosis and classification of hepatocellular carcinoma,and the difference in size can change clinical treatment decisions and follow-up methods of patients.In order to reduce the variability of size measurement,LI-RADS v2018 suggests that the size of liver lesions should not be measured on arterial phase images or DWI when the boundary of other phases is clear.However,LI-RADS didn’t make further concrete suggestions and recommendations on the phase selection of size measurement.In addition,LI-RADS stipulates that the capsule should be included when measuring the size of liver lesions.The capsule may be present in the portal venous phase or（and）hepatobiliary phase.The influence of capsule appearance in different phases on the size measurement is unclear.In this thesis,the consistency of liver lesion size measured in different phases of Gd-EOB-DTPA dynamic enhanced MRI and the influence of capsule on liver lesion size measurement were investigated.Materials and methods:Two radiologists measured the maximum diameter of 168 liver lesions confirmed by pathology in 2018 LI-RADS high-risk population on enhanced T₁ WI,enhanced T₂ WI,portal venous phase images and hepatobiliary phase images.Subgroup analysis was performed according to the unenhanced T₁ WI signal intensity（isointensity or hyperintensity,hypointensity）and the pathological results（benign and malignant lesions）.Friedman test was used to compare the difference between liver lesion size measured in different phases,and Bonferroni multiple test was used for correction.The difference in the size of liver lesions with or without capsule was analyzed by Wilcoxon signed rank test in portal venous phase images and hepatobiliary phase images.Results:For all cases,there was no statistical difference between the liver lesion size of portal venous phase images and hepatobiliary phase images.However,the size of liver lesions measured on enhanced T₁ WI and enhanced T₂ WI was significantly smaller than that of portal venous phase images and hepatobiliary phase images,and enhanced T₂ WI was significantly smaller than enhanced T₁ WI.The results of malignant lesions were basically the same as those of all cases,but the only difference was that there was no difference in the size of liver lesions measured by enhanced T₁ WI and portal venous phase images.Among benign lesions,the size of liver lesions on enhanced T₂ WI was significantly smaller than that in portal venous phase images and hepatobiliary phase images,but there was no statistical difference among enhanced T₁ WI,portal venous phase images and hepatobiliary phase images.The capsule of 62 liver lesions didn’t appear in portal venous phase and hepatobiliary phase,6 only appeared in hepatobiliary phase,69 only appeared in portal venous phase,and 31 both appeared in portal venous phase and hepatobiliary phase.Although the average rank of the phase with capsule was larger than that without capsule,there was no significant difference in the size of liver lesions in portal venous phase and hepatobiliary phase among the four groups.Conclusion:According to the recommendations of LI-RADS v2018 and our results,the size of the liver lesion should be measured in the late enhanced phase except the arterial phase.When the boundary of liver lesions is unclear or there is no post-arterial phase enhancement phase,the size can be measured on unenhanced T₁ WI,but it is not recommended to measure the size of liver lesions in unenhanced T₂ WI.The capsule should be included in the measurement of the size,and the appearance of the capsule in different phases has no effect on the size measurement.