College of Respiratory and Critical Care Medicine, Eighth Medical Center, Chinese People's Liberation Army General Hospital, Beijing, China.
Urology Surgery Ward 1, Peking Union Medical College Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, China.
Ren Fail. 2023 Dec;45(1):2194433. doi: 10.1080/0886022X.2023.2194433.
To explore a machine learning model for the early prediction of acute kidney injury (AKI) and to screen the related factors affecting new-onset AKI in the ICU.
A retrospective analysis was performed used the MIMIC-III data source. New onset of AKI defined based on the serum creatinine changed. We included 19 variables for AKI assessment using four machine learning models: support vector machines, logistic regression, and random forest. and XGBoost, using accuracy, specificity, precision, recall, F1 score, and AUROC (area under the ROC curve) to evaluate model performance. The four models predicted new-onset AKI 3-6-9-12 h ahead. The SHapley Additive exPlanation (SHAP) value is used to evaluate the feature importance of the model.
We finally respectively extracted 1130 AKI patients and non-AKI patients from the MIMIC-III database. With the extension of the early warning time, the prediction performance of each model showed a downward trend, but the relative performance was consistent. The prediction performance comparison of the four models showed that the XGBoost model performed the best in all evaluation indicators in all the time point at new-onset AKI 3-6-9-12 h ahead (accuracy 0.809 vs 0.78 vs 0.744 vs 0.741, specificity 0.856 vs 0.826 vs 0.797 vs 0.787, precision 0.842 vs 0.81 vs 0.775 vs 0.766, recall 0.759 vs 0.734 vs 0.692 vs 0.694, Fl score 0.799 vs 0.769 vs 0.731 vs 0.729, AUROC 0.892 vs 0.857 vs 0.827 vs 0.818). In the prediction of AKI 6, 9 and 12 h ahead, the importance of creatinine, platelets, and height was the most important based on SHapley.
The machine learning model described in this study can predict AKI 3-6-9-12 h before the new-onset of AKI in ICU. In particular, platelet plays an important role.
探索用于急性肾损伤（AKI）早期预测的机器学习模型，并筛选影响 ICU 新发 AKI 的相关因素。
对 MIMIC-III 数据源进行回顾性分析。根据血清肌酐的变化定义 AKI 的新发。我们使用四种机器学习模型（支持向量机、逻辑回归、随机森林和 XGBoost）对 AKI 评估纳入 19 个变量，使用准确性、特异性、精度、召回率、F1 评分和 AUROC（ROC 曲线下面积）来评估模型性能。四个模型分别预测新发生 AKI 的 3-6-9-12 h。使用 SHapley Additive exPlanation（SHAP）值评估模型的特征重要性。
我们最终分别从 MIMIC-III 数据库中提取了 1130 例 AKI 患者和非 AKI 患者。随着预警时间的延长，各模型的预测性能呈下降趋势，但相对性能保持一致。四种模型的预测性能比较表明，XGBoost 模型在新发生 AKI 的所有评估指标中表现最佳 3-6-9-12 h （准确性 0.809 对 0.78 对 0.744 对 0.741，特异性 0.856 对 0.826 对 0.797 对 0.787，精度 0.842 对 0.81 对 0.775 对 0.766，召回率 0.759 对 0.734 对 0.692 对 0.694，Fl 评分 0.799 对 0.769 对 0.731 对 0.729，AUROC 0.892 对 0.857 对 0.827 对 0.818）。在预测 AKI 6、9 和 12 h 时，基于 SHapley，肌酐、血小板和身高的重要性最为重要。
本研究描述的机器学习模型可预测 ICU 新发 AKI 前 3-6-9-12 h 的 AKI。特别的是，血小板起着重要作用。
