基于随机森林构建胃癌风险预测模型
打开文本图片集
Construction of a risk prediction model for gastric cancer based on random forest WANG Qingqing'，WU Wenbol*，WAN Shaoping'，ZHANG Linglin²，LI Yuting3，RONG Lilou1 1.Sichuan CancerHospital，Sichuan Cancer Institute，Sichuan Cancer Center，Universityof Electronic Science and Technology of China，Sichuan 610041 China;2.Renshou County Center for Disease Control and Prevention; 3.Sichuan Provincial Mental Health Center（Mianyang Third People's Hospital）
*Corresponding Author WU Wenbo，E-mail： wuwenbo1@ scszlyy.org.cn
AbstractObjective：ToconstructariskpredictionmodelforgastriccancerinSichuanprovinceandtoidentifyitspredictivefactors. Methods：Atotalof378patientsdiagnosedwithgastriccancerinfourmedicalistutionsinSichuanprovincefromJuly2O21toAugust 2022 wereselectedasthecasegroup.Atotalof378healthyindividualsudergoing physicalexaminationsduringthesameperiodwere selectedas thecontrolgroup.Therandomforestmethodwasused toconstructariskpredictionmodelforgastriccancer.Theperformance of the modelwasevaluatedbyusingaccuracyensitivity，specifityreaunder teOCcurveAUC），ibrationcurve，ndcisio curveanalysis（CA）Resutsducatioalevelccupainhiteatitak，roceedeatitakeorseainitake，kleda sundriedfoditake，etarytaste，etaryoldessndeattigsd，reakfastsatus，mokigsatus，loholosutiota drinkingstatustplirocrisicrtelalto personlitywereindependentinfluencingfactorsforgastriccance.Thetopthreeinfluencingfactorsintemsofimportancescorewere occupation，education level，and smoking status.The accuracy of the prediction model was 99.8% .The sensitivity was 100.0% .The specificity was 99.6% .The AUCwas0.9990 95% CI 0.998-1.Ooo）. The calibration curve of the prediction model indicated good consistencybetweethemodelandtheactualobservationresults.TheDCAcurvesugestedthatemodelhasgoodclinicaleficacyin preditinggastriccancer.Conclusions：TheriskpredictionmodelforgastriccancerofSichuanprovinceconstructedbasedontherandom forestmetodintissudyasgoodpredictieperformanceandiselpfulfrearlyidentificatioofhighskpopulationfogastricane. Keywordsgastric cancer;random forest; risk prediction model; influencing factors
2022年，我国胃癌新发病例数35.8万例，死亡26.0万例，占全球胃癌新发病例的 37.0% ，占全球死亡病例的 39.4%[1] ，均处于较高水平。（剩余9521字）