乳腺癌术后化疗病人感染风险预测模型的 构建

打开文本图片集

Construction of an infection risk prediction model for breast cancer patients undergoing postoperative chemotherapy

ZHONG Yanlan，ZHANG Qing，PENG Yun

GanzhouPeople'sHospital，Jiangxi341oo0China

CorrespondingAuthor ZHONGYanlan，E-mail： zhongyanlanl@163.com

AbstractObjective：Toconstructaninfectionriskprediction model forbreastcancerpatientsundergoingpostoperativechemotherapy. Methods：Atotalof368patientswhounderwentpostoperativechemotherapyforbreastcancerandadmitedtothethyroidandbreast surgerydepartmentofGanzhouPeople'sHospitalfromOctober2O2OtoJune2023 wereselectedas thestudysubjects.BasedonLogistic regreionnalss，aifcatiogreiontreendackpropagaralrkagiskpdictioodelsfectioi breastcancerpatientsundergoingpostoperativechemotherapywereconstructed.The predictivevaluewas analyzed bycomparingthe receiveroperatingcharacteristccurvesof theprediction models.Results：Infectionsocured i62breastcancerpatientsundergoing postoperativechemotherapy，primarilyafectingtherespiratorytract.MultivariateLogistcregressionanalysissultsshowedatbone marowsuppressonCeactiprotedprocalioereepdentiuecingfactosforfetisieascit undergoing postoperative chemotherapy （P<0.05） .The classification regression tree model showed that C-reactive protein，procalcitonin， drainagedurationanddiabeteswereinfluencingfactorsforinfections.Thebackpropagationneuralnetwork modelshowedthatthe importanceoffactorsafectinginfectionsinbreastcancerpatientsundergoingpostoperativechemotherapywasrankedasfollows： C-reactive protein>procalcitonin>combined diabetes>length of hospital stay > bone marrow suppression>drainage duration>serum albumin>chemotherapycycles.Among thethreemodels，thebackpropagationeuralnetworkmodeldemonstratedthebestpredictive performance.TheareaunderthereceiveroperatingcharacteristiccurvewasO96.Thesensitivitywas1.Oo.ThespeificitywasO.931.

Conclusions：TeinfluencingfactorsofinfectionriskinbreastcancerpatientsundergoingpostoperativechemotherapyincludeCreactie protein，proalionbtes，gthfsitalaydoapiotcoiskictoodet cancerpatientsundergoingpostoperativechemotherapyconstructedbasedonmachinelearmingalgorithmsallexhibitedgood performance，with the back propagation neural network model demonstrating the best predictive performance.

Keywordsc;soatieaditidel;gistalis regression tree; back propagation neural network; influencing factor

摘要目的：构建乳腺癌术后化疗病人感染风险预测模型。（剩余10765字）