摘要
目的:基于粒子群优化算法(PSO)的XGBoost模型(PSO-XGBoost)预测中老年患者经皮肾镜碎石术(PCNL)后的泌尿系统结石残留情况。方法:选取2014年1月至2018年12月在某三级甲等医院接受PCNL治疗的596例中老年肾结石患者为研究对象。将数据集按7:3的比例随机分为训练集和测试集,建立PSO-XGBoost模型,以准确度、精准度、召回率、F1值和ROC曲线下面积(AUC)等指标评估模型性能。结果:PSO-XGBoost模型预测PCNL术后泌尿系统结石残留的准确度、精准度、召回率、F1值和AUC均优于支持向量机、近邻算法、决策树和BP神经网络模型。结论:PSO-XGBoost模型可更快、更准确地预测中老年患者PCNL术后是否有泌尿系统结石残留,为制定术后个性化治疗、护理方案提供参考。
Objective To predict the kidney stone residue in middle-aged and elderly patients after percutaneous nephrolithotomy(PCNL)by the XGBoost model(PSO-XGBoost)based on particle swarm optimization algorithm(PSO).Methods A total of 596 middle-aged and elderly patients with kidney stone who received PCNL at a Grade-A Tertiary hospital from January 2014 to December 2018 were selected for this study.The dataset was randomly divided into the training set and the test set at a ratio of 7:3,the PSO-XGBoost model was established,and the performance was evaluated by accuracy,precision,recall,F1 score,and AUC(area under the curve).Results The accuracy rate,precision rate,recall rate,F1 score and AUC of PSO-XGBoost model outperformed those of other models,including support vector machine,nearest neighbor algorithm,decision tree and BP neural network model.Conclusion The PSOXGBoost model can predict the presence of kidney stone residue after PCNL in middle-aged and elderly patients more quickly and accurately,which can provide reference for making personalized treatment and nursing plans after PCNL.
作者
唐圣晟
刘泓泽
李琳
罗云汉
廖芝美
周毅
TANG Shengsheng;LIU Hongze;LI Lin;LUO Yunhan;LIAO Zhimei;ZHOU Yi(Zhongshan School of Medicine,Sun Yat-sen University,Guangzhou 510080,Guangdong Province,China)
出处
《中国数字医学》
2023年第8期24-29,共6页
China Digital Medicine
基金
国家重点研发计划(2022YFC3601600,2021YFC2009400)
广州市科技计划项目(202206010028)
广东省科技创新战略专项(202011020004)
2023大学生创新创业训练项目(202310657)。
作者简介
通信作者:周毅,Email:zhouyi@mail.sysu.edu.cn。
