基于数据驱动的中青年糖尿病前期病人转化风险预测模型的构建

打开文本图片集

AbstractObjective：Toconstructatransformationriskpredictionmodelforyoungandmiddleaged prediabeticpatientsbasedonthe datafromthe health informationplatformofYinzhou districtof NingbocityZhejiang province.Methods：Aretrospectivecohort study design waadopted.Combined withliteraturereviewandexpertrecommendations，24predictorswereincluded.LASSOregressionwas aplied forvarablesreening.ThepredictionmodelwasconstructedbyusingLgisticregressonandfourmachineleaingalgorithms. Results：A total of7 379 patientswere included，among whom 731 cases （9.91% ）transformed into type2diabetes.Therandom forest modelperformedrelativelywellIthadagooddiscriminatoryability.However，itsaccracywasmoderate.Conclusions：The transformationrisk prediction modelfor young andmiddle-aged prediabetic patients basedonrandom forests has gooddiscrimnatory ability.Itcould provideareferenceforoptimizing the precise managementofyoungand middle-aged prediabeticpatients.

Keywordsmidegdndounpeope;prdabeestdeteseliusaceang;prdictieodel;iueingfct;usig

摘要目的：基于浙江省宁波市鄞州区健康信息平台数据，构建中青年糖尿病前期向2型糖尿病转化风险预测模型。（剩余14722字）