新型关联规则模型对PSA灰区前列腺癌的预测价值

打开文本图片集
[Abstract]Objective:BycreatinganewassociationrulemodelbasedonclinicaldatasuchasProstate ImagingReporting andDataSystemversion2.1(PI-RADSV2.1)score,toevaluatetheeficiencyof themodelforpredictingprostatecancer(PCa) intheprostate-specificantigen(PSA)greyzone.Methods:Atotalof1OoOpatientswithprostatediseasewereretrospectively colected.Thepatientswererandomlydividedintoamodelinggroup(70cases)andavalidationgroup(30Ocases).The asociationruleanalysiswasperformedontheclinicaldataofthemodelinggroup,takingtheclinicaldatasuchasPI-RADS V2.1scoreastheantecedentandCaastheconsequent,teefectivestrongasociationruleswerecalculatedandthepredictive model wascreated.Thecalibrationperformancewasasssedusing thecalibrationcurve.Internalvalidationwasalso performed.ROCcurvewasusedtoanalyzetheAUCofthemodelforpredictingPCa.Externalvalidationwasalsoperformed. Results:Thepredictionresultsof thenewasociationrule modelshowedthattheincidenceof PCawas79%whenPI-RADS V2.1 score ≥3points,age≥7Oyears,prostatevolume<50mL,andtotalPSA(tPSA)≥7ng/mL.Theinternal validationshowed thattheconcordanceindex(C-index)of the modelwasO.762(95%CI0.639-0.810),andtherewasagoodconsistency betweenthepredictedandobservedvaluesonthecalibrationcurve.TheAUCofthemodelforpredictingPCawas0.893, which washigherthanthePI-RADSV2.1score(0.831),age(0.686),prostatevolume(0.722),and tPSA(0.634).Theexternal validationshowedthattheAUCofthemodelforpredictingPCainthevalidationgroupwasO.884.Theresultwasingod agreement with theROC curveof the modeling group,and thedifference was not statistically significant ( χ∠ =8.098,P>0.05). ThepredictionperformanceofthemodelforPCaindiferentprostateregionscouldmettheclinicalrequirements.Conclusions: Thenewassociationrulemodelbasedon clinicaldatasuch asPI-RADSV2.1score shows favorableperformance in predicting PCa and has potential clinical application value.
[Keywords]Prostate Imaging ReportingandData Systemversion2.1;Assciationruleanalysis;Prostatespecificantigen; Greyarea;Prostatecancer;Forecast
前列腺癌是发生于前列腺上皮的恶性肿瘤,是男性泌尿生殖系统最常见的恶性肿瘤[1-3]。(剩余8462字)