论文部分内容阅读
Objective This studyaimedtoevaluatetheutilityofgadoxeticacid-enhancedmagneticresonanceimaging(MRI)forthequantitativeevaluationofearlyhepatocellularcarcinoma(HCC)andbenignhepatocellularnodulesinthehepatobiliaryphase(HBP)inpatientswithcirrhosis. Methods Thesubjectscomprised32patientswith49lesions(31earlyHCCsand18benignhepatocellularnodules)whohadundergonegadoxeticacid-enhanced3.0TMRI,includingbaselinesequences.Thebenignhepatocellularnoduleswereconfirmedbypathologicalexamination,orwithimagingevidenceifthelesioneitherdisappearedcompletelyorshowednochangeinthe follow-upimagesforatleast12monthsafteritsinitialappearance.HCCwasconfirmedatsurgicalresection,typicalimagingfindings(markedarterialenhancementandfastwashout),orbyshowingintervalgrowthwithtypicalenhancementpatternsat follow-upMRI.Chemical-shift-selectivefat-suppressedT1-weightedsequenceswereobtainedbeforeand20minaftertheinjectionofgadoxeticacid.Thesignalintensity(SI)ratiowasmeasuredastheSIofaHCCorabenignlesiondividedbytheSIoftheadjacentlivertissue.FisherexacttestwasusedtocomparethefindingsonT2-weightedimages,DWIandHBPimagesofthetwogroups.ThedifferencesofmeansizesandHBPSIratiosbetweenthetwogroupswerecomparedwiththeMann-WhitneyU test. Results Ongadoxeticacid–enhancedHBPimages,84%ofbenign(15of18)lesionsand97%HCCs(30of31)werehypointensecomparedwithsurroundinglivertissue(p=0.134).ThemeanHBPSIratioofHCCs(0.56±0.13[standarddeviation])was significantlylower(P<0.001)thanthatofbenignlesions(0.81±0.27).TheoptimalcutoffvalueofHBPSIratiofordistinguishing betweenHCCsandbenignhepatocellularlesionswas0.58. Conclusion TherearestillapartofbenignhepatocellularnodulesshowinghypointensityintheHBPofgadoxeticacid-enhancedMRI.QuantitativemeasurementsofHBPSIratioscouldbeusedtodistinguishbenignhepatocellularnodulesfromearlyHCCs;aHBPSIratiobelow0.58andvisibilityonDWIandT2-weightedimageswerefindingsspecificforHCCsratherthanbenignhepatocellula nodules.