Division of Cardiac Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania.
Auton Lab, Carnegie Mellon University, Pittsburgh, Pennsylvania.
Ann Thorac Surg. 2020 Jun;109(6):1811-1819. doi: 10.1016/j.athoracsur.2019.09.049. Epub 2019 Nov 7.
This study evaluated the predictive utility of a machine learning algorithm in estimating operative mortality risk in cardiac surgery.
Index adult cardiac operations performed between 2011 and 2017 at a single institution were included. The primary outcome was operative mortality. Extreme gradient boosting (XGBoost) models were developed and evaluated using 10-fold cross-validation with 1000-replication bootstrapping. Model performance was assessed using multiple measures including precision, recall, calibration plots, area under the receiver-operating characteristic curve (C-index), accuracy, and F1 score.
A total of 11,190 patients were included (7048 isolated coronary artery bypass grafting [CABG], 2507 isolated valves, and 1635 CABG plus valves). The Society of Thoracic Surgeons Predicted Risk of Mortality (STS PROM) was 3.2% ± 5.0%. Actual operative mortality was 2.8%. There was moderate correlation (r = 0.652) in predicted risk between XGBoost and STS PROM for the overall cohort and weak correlation (r = 0.473) in predicted risk between the models specifically in patients with operative mortality. XGBoost demonstrated improvements in all measures of model performance when compared with the STS PROM in the validation cohorts: mean average precision (0.221 XGBoost vs 0.180 STS PROM), C-index (0.808 XGBoost vs 0.795 STS PROM), calibration (mean observed-to-expected mortality: XGBoost 0.993 vs 0.956 STS PROM), accuracy (1%-3% improvement across discriminatory thresholds of 3%-10% risk), and F1 score (0.281 XGBoost vs 0.230 STS PROM).
Machine learning algorithms such as XGBoost have promise in predictive analytics in cardiac surgery. The modest improvements in model performance demonstrated in the current study warrant further validation in larger cohorts of patients.
本研究评估了机器学习算法在预测心脏手术手术死亡率中的预测能力。
纳入了 2011 年至 2017 年在一家单机构进行的索引成人心脏手术。主要结局是手术死亡率。使用 10 折交叉验证和 1000 次重复引导进行了极端梯度增强（XGBoost）模型的开发和评估。使用多种指标评估模型性能，包括精度、召回率、校准图、接收者操作特征曲线（C 指数）下的面积、准确性和 F1 分数。
共纳入 11190 例患者（7048 例单纯冠状动脉旁路移植术[CABG]、2507 例单纯瓣膜手术和 1635 例 CABG 加瓣膜手术）。胸外科医生协会预测死亡率（STS PROM）为 3.2%±5.0%。实际手术死亡率为 2.8%。在整个队列中，XGBoost 与 STS PROM 的预测风险之间存在中度相关性（r=0.652），而在具有手术死亡率的患者中，两种模型之间的预测风险相关性较弱（r=0.473）。与 STS PROM 相比，XGBoost 在验证队列中的所有模型性能指标上均有所提高：平均平均精度（0.221 XGBoost 比 0.180 STS PROM）、C 指数（0.808 XGBoost 比 0.795 STS PROM）、校准（观察到的预期死亡率平均值：XGBoost 0.993 比 STS PROM 0.956）、准确性（在 3%-10%风险的歧视性阈值上提高 1%-3%）和 F1 分数（0.281 XGBoost 比 0.230 STS PROM）。
机器学习算法，如 XGBoost，在心脏手术预测分析中有应用前景。在本研究中，模型性能的适度提高需要在更大的患者队列中进一步验证。
