International journal of cardiology · 2025
Predicting prolonged length of in-hospital stay in patients with non-ST elevation myocardial infarction (NSTEMI) using artificial intelligence.
СтатьяПоддержка исследования: вне правительства СШАНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Razavi Seyed Reza, Szun Tyler, Zaremba Alexander C, Cheung Seth, Shah Ashish H, Moussavi Zahra
Аннотация
BACKGROUND: Patients presenting with non-ST elevation myocardial infarction (NSTEMI) are typically evaluated using coronary angiography and managed through coronary revascularization. Numerous studies have demonstrated the benefits of expedited discharge following revascularization in this patient population. However, individuals with concomitant heart failure, hemodynamic instability, or arrhythmias often necessitate prolonged hospitalization. Using aortic pressure (AP) wave assessment, we aim to predict a prolonged length of stay (> 4 days, PLoS) in patients with NSTEMI treated with percutaneous coronary intervention (PCI).
METHODS: In this single-center, retrospective cohort study, we included 497 patients with NSTEMI [66.3 ± 12.9 years, 37.6 % (187) females]. We developed a predictive model for PLoS using features primarily extracted from the AP signal recorded throughout PCI. We performed feature selection using recursive feature elimination (RFE) with cross-validation and built a machine learning (ML) model using the CatBoost tree-based classifier. The decision-making process of the ML model was analyzed using SHapley Additive exPlanations (SHAP).
RESULTS: We achieved average accuracy, specificity, sensitivity, precision, and receiver operating characteristic curve area under the curve (AUC) values of 77 %, 78 %, 76 %, 67 %, and 77 %, respectively. Using SHAP, we identified the ejection systolic period, ejection systolic time, the difference between systolic blood pressure and dicrotic notch pressure (DesP), the age modified shock index (mSI_age) and mean arterial pressure (MAP) as the most characteristic features extracted from the AP signal.
CONCLUSIONS: In conclusion, this study demonstrates the potential of using ML and features extracted from the AP signal to predict PLoS in patients with NSTEMI.
METHODS: In this single-center, retrospective cohort study, we included 497 patients with NSTEMI [66.3 ± 12.9 years, 37.6 % (187) females]. We developed a predictive model for PLoS using features primarily extracted from the AP signal recorded throughout PCI. We performed feature selection using recursive feature elimination (RFE) with cross-validation and built a machine learning (ML) model using the CatBoost tree-based classifier. The decision-making process of the ML model was analyzed using SHapley Additive exPlanations (SHAP).
RESULTS: We achieved average accuracy, specificity, sensitivity, precision, and receiver operating characteristic curve area under the curve (AUC) values of 77 %, 78 %, 76 %, 67 %, and 77 %, respectively. Using SHAP, we identified the ejection systolic period, ejection systolic time, the difference between systolic blood pressure and dicrotic notch pressure (DesP), the age modified shock index (mSI_age) and mean arterial pressure (MAP) as the most characteristic features extracted from the AP signal.
CONCLUSIONS: In conclusion, this study demonstrates the potential of using ML and features extracted from the AP signal to predict PLoS in patients with NSTEMI.