Machine Learning-Based Prediction of 30-Day All-Cause Mortality in Patients Hospitalized With Acute Pulmonary Embolism

Abstract

Patients diagnosed with acute pulmonary embolism (PE) are at an elevated risk for short-term mortality. An accurate prognostic model is important in guiding appropriate management. We aimed to use machine learning (ML) to predict 30-day all-cause mortality in patients diagnosed with acute PE. METHODS: 439 patients (48% men, 61AE15 years) diagnosed with acute PE at our institution were retrospectively analyzed. We included 101 variables from a range of domains including demographics, clinical, laboratory, echocardiographic, and CT imaging as candidate predictors. Machine learning algorithms including extreme gradient boosting (XGBoost), gradient boosting machine (GBM), random forest (RF), deep neural networks (DNN), and generalized linear models (GML) were evaluated on their classification performance and validated with 5-fold cross-validation. The PE severity index (PESI) and its simplified version (sPESI) were used as reference models. RESULTS: XGBoost was the best performing model in predicting 30-day all-cause mortality (AUC, 0.922 (95% confidence interval [

Machine Learning-Based Predictive Models for Patients with Venous Thromboembolism: A Systematic Review

“…Nevertheless, if it is not performed properly data leakage can happen. In every case an unseen part of the dataset must be kept for internal validation purposes and this was not clear in many [28][29][30]34,[38][39][40][41]51,53,54,[56][57][58]61,62,64,77,84,89,126,127 of the reviewed studies. This review highlighted a significant problem in AI-related research, namely the lack of standardized reporting methods.…”