- NEWS . 06 Apr 2020
Prediction of short-term mortality post-PCI using the decision-tree model
The decision-tree (DT) model has been shown to be effective in predicting 30-day mortality after percutaneous coronary intervention (PCI) for acute myocardial infarction (AMI).
This is based on a study that analysed data from the Taiwan National Health Insurance Research Database (NHIRD) to construct a variety of mortality prediction models. The goal was to find an appropriate multivariate predictor of post-PCI patient mortality.
The study cohort comprised 3,421 patients diagnosis with AMI undergoing PCI between 2004 and 2013. Of these, 3,079 and 342 patients were included in the training and test groups, respectively. Each patient had 22 input features and two output features that represented mortality. The prediction models implemented in the study included an artificial neural network model, a linear discriminant analysis classifier, a logistic regression model, a naive Bayes classifier, a support vector machine, as well as the DT model.
Results showed that the DT model was the most suitable in terms of performance and real-word applicability, achieving an area under receiving operating characteristic of 0.895 (95% confidence interval, 0.865–0.925), weighted-average F1 score of 0.969, precision of 0.971 and recall of 0.974.
Hsieh MH, et al. A fitting machine learning prediction model for short-term mortality following percutaneous catheterization intervention: a nationwide population-based study. Ann Transl Med 2019;7:732.