机器学习模型预测肾结石病24小时尿液异常
本研究利用电子健康记录数据(n=1314)训练XGBoost模型预测肾结石患者的24小时尿液异常,并与逻辑回归及二者集成模型比较。XGBoost对尿量、尿酸和尿钠异常预测较准,逻辑回归在尿pH和枸橼酸预测上优于XGBoost,集成模型对草酸和枸橼酸预测最准。作者认为机器学习预测尿液化学可行,进一步优化可辅助饮食或药物预防。
要点
- XGBoost预测尿量异常准确率98%、AUC-ROC 0.59,尿酸69%、0.73,尿钠升高71%、0.79。
- 逻辑回归预测尿pH异常AUC-ROC 0.66优于XGBoost的0.57,枸橼酸0.69优于0.64。
- 集成模型预测草酸异常准确率65%、AUC-ROC 0.70,枸橼酸65%、0.69,总体与单模型相似。
核心结论
- 作者认为机器学习方法预测肾结石病的尿液化学是可行的,进一步优化性能可促进饮食或药物预防。
研究设计与样本
- 研究利用电子健康记录衍生数据(n=1314)训练机器学习模型(XGBoost)预测24小时尿液异常,并与逻辑回归模型及二者集成模型进行比较。
主要发现
- XGBoost模型区分24小时尿液异常的性能尚可,与逻辑回归相当。
- XGBoost模型预测尿量异常准确率98%、AUC-ROC 0.59,尿酸异常准确率69%、AUC-ROC 0.73,尿钠升高准确率71%、AUC-ROC 0.79。
- 逻辑回归模型预测尿pH异常的AUC-ROC为0.66,优于XGBoost的0.57;预测枸橼酸异常的AUC-ROC为0.69,优于XGBoost的0.64。
- 集成模型预测草酸异常准确率65%、AUC-ROC 0.70,枸橼酸异常准确率65%、AUC-ROC 0.69,总体预测性能与XGBoost或逻辑回归单模型相似。
- 体重指数、年龄和性别是训练所有结局模型时最重要的三个特征。
临床怎么用
- 原文未据此给出推荐。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVE: To help guide empiric therapy for kidney stone disease, we sought to demonstrate the feasibility of predicting 24-hour urine abnormalities using machine learning methods. METHODS: We trained a machine learning model (XGBoost [XG]) to predict 24-hour urine abnormalities from electronic health record-derived data (n = 1314). The machine learning model was compared to a logistic regression model [LR]. Additionally, an ensemble (EN) model combining both XG and LR models was evaluated as well. Models predicted binary 24-hour urine values for volume, sodium, oxalate, calcium, uric acid, and citrate; as well as a multiclass prediction of pH. We evaluated performance using area under the receiver operating curve (AUC-ROC) and identified predictors for each model. RESULTS: The XG model was able to discriminate 24-hour urine abnormalities with fair performance, comparable to LR. The XG model most accurately predicted abnormalities of urine volume (accuracy = 98%, AUC-ROC = 0.59), uric acid (69%, 0.73) and elevated urine sodium (71%, 0.79). The LR model outperformed the XG model alone in prediction of abnormalities of urinary pH (AUC-ROC of 0.66 vs 0.57) and citrate (0.69 vs 0.64). The EN model most accurately predicted abnormalities of oxalate (accuracy = 65%, ROC-AUC = 0.70) and citrate (65%, 0.69) with overall similar predictive performance to either XG or LR alone. Body mass index, age, and gender were the three most important features for training the models for all outcomes. CONCLUSION: Urine chemistry prediction for kidney stone disease appears to be feasible with machine learning methods. Further optimization of the performance could facilitate dietary or pharmacologic prevention.
引用
- Kavoussi NL, Floyd C, Abraham A, et al. Machine Learning Models to Predict 24 Hour Urinary Abnormalities for Kidney Stone Disease. Urology. 2022. PMID: 35853510. doi:10.1016/j.urology.2022.07.008
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。