基于超声衰减参数评估超声联合临床参数模型对肝脏脂肪含量的预测价值

打开文本图片集

[Abstract]Objective：Toinvestigatethepredictivevalueofmulti-parametermodelbasedonultrasoundatenuationparameters andclinicaldataforliverfatcontent.Methods：Atotalof615peoplewhounderwenthealthexaminationwereselected， andallofthemunderwentultrasoundexaminationtomeasurethevisceralfatthickness（VFT）andsubcutaneousfat thickness（SFT），andtheatenuationcoefcient（AC）ofultrasound-guidedatenuationparameter（UGAP）wererecorded. Amongthesubjects，68casesunderwentMRIprotondensityfatfraction（PDFF）examination，andthecorelationbetween MRI-PDFFvalueandACvaluewasanalyzed.Theremaining 547subjectsinvolvedinthepredictionmodelconstruction andvalidation，wererandomlydividedintoatrainingcohort（38Ocases）andvalidationcohort（167cases）ataratioof 7：3 .Thestepwiseregression methodwasusedtoconstructthepredictionmodelfor liverfat content prediction.Pearson correlationandROCcurvewasusedforanalyzethepredictiveeffciency.Results：UGAP-ACvaluewashighlycorrelated with MRI-PDFF value（ r=0.931 ）.Ultimately，the fatty liver degree diagnosed by ultrasound，VFT，SFT and triglycerides were incorporatedintothepredictivemodel.Thefinalconstructedmultiplelinearregresionequationwas：Logit（ACpredicted value） = 0.427+ 0.082× fatty liver degree + 0.001×SFT+0.000388×VFT+0.003× triglycerides.The predicted valueof the model washighlycorrelated with UGAP-AC value（ r=0.906 ）.ROC curve showed that the model had the optimal eficacyindiagnosinghepaticsteatosisinthevalidationcohortwhenthecut-offvaluewasO.57，withanAUCof0.985，a sensivityof1.OOandaspecfcityofO.91.Conclusions：UGAPenablesquantitativeanalysisofliverfat.Thepredictive valueofthemodelintegratingabdominalultrasoundandclinicaldatademonstratesastrongcorelationwith UGAP-AC， whichcanbeusedforquantitativeestimationofliverfatcontentandidentificationofpopulationswithhepaticsteatosis.This model exhibits favorable clinical utility and excellent diagnostic performance.

[KeyWords]Quantificationofliverfat；Fattydegenerationofliver；Multiplelinearregresionmodel；Ultrasound-guided attenuation parameter;Ultrasonography;Serology

肝脏脂肪异常沉积超过肝脏总重量的 5% 为肝脂肪变性，患病率约 29.2% ，且呈逐年上升趋势，可引起肝脏代谢功能障碍，增加代谢相关脂肪性肝病、肝硬化、肝细胞癌患病风险[1]。（剩余7561字）