The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel; Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel.
The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel; Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel; Adult Bone Marrow Transplantation Service, Memorial Sloan Kettering Cancer Center, New York, NY, United States of America.
J Cardiol. 2022 Mar;79(3):342-351. doi: 10.1016/j.jjcc.2021.11.006. Epub 2021 Nov 29.
Clinical risk assessment with quantitative formal risk scores may add to intuitive physician risk assessment and are advised by the international guidelines for the management of acute coronary syndrome (ACS) patients. Most previous studies have used the binary regression/classification approach (dead/alive) for long-term mortality post-ACS, without considering the time-to-event as in survival analysis. The use of machine learning (ML)-based survival models has yet to be validated. The primary objective was to compare survival prediction performance of 1-year mortality following ACS of two newly developed ML-based models [random survival forest (RSF) and deep learning (DeepSurv)] with the traditional Cox-proportional hazard (CPH) model. The secondary objective was external validation of the findings.
This was a retrospective, supervised learning data mining study based on the Acute Coronary Syndrome Israeli Survey (ACSIS) and the Myocardial Ischemia National Audit Project (MINAP). The ACSIS data were divided to train/test in a 70/30 fashion. Next, the models were externally validated on the MINAP data. Harrell's C-index, inverse probability of censoring weighting (IPCW), and the Brier-score were used for models' performance comparison.
RSF performed best among the three models, with Harrell's C-index on training and testing sets reaching 0.953 and 0.924 respectively, followed by CPH multivariate selected model (0.805/0.849), CPH Univariate selected model (0.828/0.806), DeepSurv model (0.801/0.804), and the traditional CPH model (0.826/0.738). The RSF model also had the highest performance on the validation data set with 0.811 for Harrell's C-index, 0.844 for IPCW, and 0.093 for Brier score. The CPH model performance on the validation set had C-index range between 0.689 to 0.790, 0.713 to 0.826 for IPCW, and 0.094 to 0.103 Brier score.
RSF survival predictions for long-term mortality post-ACS show improved model performance compared with the classic statistical method. This may benefit patients by allowing better risk stratification and tailored therapy, however further prospective evaluations are required.
使用定量形式风险评分进行临床风险评估可能会补充直观的医生风险评估，并为急性冠状动脉综合征（ACS）患者管理的国际指南所建议。大多数先前的研究都使用二元回归/分类方法（存活/死亡）来预测 ACS 后的长期死亡率，而没有考虑生存分析中的时间事件。基于机器学习（ML）的生存模型的使用尚未得到验证。主要目的是比较两种新开发的基于 ML 的模型[随机生存森林（RSF）和深度学习（DeepSurv）]与传统 Cox 比例风险（CPH）模型对 ACS 后 1 年死亡率的生存预测性能。次要目标是对研究结果进行外部验证。
这是一项基于急性冠状动脉综合征以色列调查（ACSIS）和心肌缺血国家审计项目（MINAP）的回顾性、有监督的学习数据挖掘研究。ACSIS 数据以 70/30 的比例分为训练/测试。接下来，将模型在 MINAP 数据上进行外部验证。使用 Harrell 的 C 指数、逆概率 censoring 加权（IPCW）和 Brier 评分来比较模型的性能。
在这三种模型中，RSF 的表现最佳，训练集和测试集的 Harrell 的 C 指数分别达到 0.953 和 0.924，其次是 CPH 多变量选择模型（0.805/0.849）、CPH 单变量选择模型（0.828/0.806）、DeepSurv 模型（0.801/0.804）和传统的 CPH 模型（0.826/0.738）。RSF 模型在验证数据集上也具有最高的性能，Harrell 的 C 指数为 0.811，IPCW 为 0.844，Brier 分数为 0.093。CPH 模型在验证集上的性能 C 指数范围在 0.689 到 0.790 之间，IPCW 在 0.713 到 0.826 之间，Brier 分数在 0.094 到 0.103 之间。
与经典统计方法相比，RSF 对 ACS 后长期死亡率的生存预测显示出了改进的模型性能。这可能通过允许更好的风险分层和量身定制的治疗来使患者受益，但是需要进一步的前瞻性评估。
