The LightGBM model based on lipid-related biomarkers predicted acute coronary syndrome with an AUC of 0.790 in the test set of hospitalized Chinese patients.
Observational (n=10,127)
Yes
Does a machine learning-derived risk score based on lipid-related biomarkers accurately predict acute coronary syndrome in hospitalized patients?
A novel machine learning-derived risk score incorporating lipid-related biomarkers (UHR, TyG, Lp(a)) and clinical features can effectively stratify the risk of acute coronary syndrome in hospitalized patients.
Effect estimate: AUC 0.790 in test set for LightGBM model
Absolute Event Rate: 15.2% vs 15.8%
Background: This study aimed to develop and test an explainable machine learning (ML) predictive model based on lipid-related biomarkers to predict acute coronary syndrome (ACS) in hospitalized patients. Methods: A total of 10,127 consecutive hospitalized patients at three large hospitals were retrospectively studied between 2022 and 2024. ACS incidence was recorded as the primary outcome. Eight ML models were used to calculate the risk of ACS during hospitalization and to distribute patients into low-, intermediate-, and high-risk groups. Results: All patients were randomly divided into a 70% training set (n = 7088) and a 30% test set (n = 3039). ACS occurred in 1119 (15.8%) and 461 (15.2%) patients, respectively. The Light Gradient Boosting Machine (LightGBM) exhibited the best predictive performance (area under the curve, 0.829) for ACS in the training set. The final model, which included the top 10 features from the LightGBM model, including lipid-related markers and clinical features, achieved a C-index of 0.781 on the test set and demonstrated a significant ability to stratify patients into low-, intermediate-, and high-risk groups. Conclusion: We constructed a risk-stratification model based on lipid-related biomarkers derived from ML models to predict ACS in hospitalized patients, which could assist in identifying patients with high discriminatory capacity.
Wan et al. (Wed,) conducted a observational in Hospitalized adult Chinese Han patients without advanced heart failure (NYHA III-IV), chronic coronary syndrome, major surgery or infection in preceding months, and severe systemic diseases affecting lipid metabolism (n=10,127). Light Gradient Boosting Machine (ML risk prediction model based on lipid-related biomarkers and clinical features) vs. Other machine learning models and logistic regression model (Decision Tree, Random Forest, XGBoost, SVM, MLP, Elastic Net, Logistic Regression) was evaluated on Incidence of acute coronary syndrome (ACS) during hospitalization diagnosed per 2014 ACC/AHA guidelines (AUC 0.790 in test set for LightGBM model). The LightGBM model based on lipid-related biomarkers predicted acute coronary syndrome with an AUC of 0.790 in the test set of hospitalized Chinese patients.