The Clinical Application Of Liver Ultrasound Imaging Reporting And Date System Classification Method Of Liver Occupying Lesions

Part One Application of Ultrasound-based Liver Imaging Reporting and Data System Categorization for Diagnosis of Focal Liver LesionsObjective To evaluate the application of ultrasound-based liver imaging reporting and data system(LIRADS-US)in the diagnosis of focal liver lesions(FFL).Methods This retrospective study included the 429 liver focal lesions in 377 patients for which pathologic diagnosis or follow-up imaging findings including CT or MRI were available.Multivariate logistic regression analysis and χ2 test were performed to independent predictors of malignancy from features that could have discriminatory value in distinguishing benign from malignant lesions: shape,margins,mass effect,echogenicity,compressibility,background,metastatic signs,nodule in nodule sign,flow/Doppler,and follow-up information.Using these suspicious features in LIRADS-US,all the nodules were stratified.The diagnostic efficacy was assessed by sensitivity,specificity,negative predictive value(NPV),positive predictive value(PPV),and receiver operating characteristic(ROC)curves obtained respectively in LIRADS-US category 3 and 4a.Results The final ten independent predictors of malignant focal liver lesions were irregular shape,nodule in nodule sign,ill-defined,lobulated or irregular margins,extrahepatic extension,halo,edge blood-vessel sign,vascular disruption and bypass,or an artery blood flow spectrum,lack of compressibility,nodular maximal diameter more than 1cm in the background of cirrhosis,presence of metastasis,diameter of a tumor increased more than 3mm in a three months follow-up.Odds ratios(OR)of these features were 3.704,6.834,7.152,3.237,6.896,4.913,5.632,6.509,8.276,and 4.70,respectively.As the number of suspicious ultrasound features increased,the risk of malignancy also increased.If LIRADS-US category 4a is defined as benign and category 4b,4c and 5 are as malignant,the sensitivity,specificity,NPV,PPV,accuracy,and area under the curve(AUC)are 85.41%,95.80%,85.41%,95.80%,93.47%,0.906,respectively.However,When LIRADS-US category 3 was defined as benign and category 4,5 were as malignant,the sensitivity,specificity,NPV,PPV,accuracy,and AUC were 91.67%,77.18%,53.66%,96.98%,80.41%,0.844,respectively.The evaluation of two physicians showed good consistency(Kappa=0.812,P<0.05).Conclusion LI-RADS classification can effectively diagnose focal liver lesionPart Two LI-RADS-US with Contrast-enhanced Ultrasound in Identification of Focal Liver Lesions.Objective To develop a contrast-enhanced ultrasound algorithm(LI-RADS-CEUS ?=ultrasound-basedliver ?imaging reporting and data system with contrast-enhanced ultrasound)in the differential diagnosis of benign and malignant focal liver lesions.Method A CEUS algorithm(LI-RADS-CEUS)was designed analogously to LI-RADS-US.LI?-RADS-CEUS was evaluated in 100 patients with 100 FLLs categorized 3/4,using LI-RADS-US criteria.Categorized 100 FLLs according to enhancement patterns.Results were compared with pathology as reference golden standards.The sensitivity,specificity,PPV,NPV,accuracy and AUC of ROC were compared between LI-RADS-CEUS and LI-RADS-US.Results The Sensitivity,specificity,NPV,PPV,accuracy,and AUC of LI-RADS-US were 80.56%,82.80%,72.50%,88.33%,82.00% and 0.817,respectively.While the same statistical indicators of LI-RADS-CEUS were 91.67%,93.75%,89.19%,95.24%,93% and 0.927.The Sensitivity,specificity and accuracy of LI-RADS-CEUS were superior to those of LI-RADS-US(P>0.05).Conclusion LI-RADS-CEUS is more effective in the differential diagnosis of benign and malignant focal liver lesions.