BD - Earth day 2024

Machine Learning-based Models for Prediction of the Risk of Stroke in Coronary Artery Disease Patients Receiving Coronary Revascularization

Lulu Lin, Li Ding, Zhongguo Fu, Lijiao Zhang

To construct several prediction models for the risk of stroke in coronary artery disease (CAD) patients receiving coronary revascularization based on machine learning methods.

Coronary artery disease (CAD) is the most common cardiovascular diseases wherein atherosclerosis occurs in one or more of the coronary arteries. CAD was reported to be one of the major causes of mortality in both the developed and developing countries. Currently, percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) are common coronary revascularization procedures. With the development and application of drug-eluting stents and minimally invasive surgery, the prognosis of patients undergoing PCI or CABG was improved, but some patients still have postoperative adverse cardiovascular events, which result in worse prognosis. Stroke is a cerebrovascular disorder which is the second leading cause of mortality and morbidity worldwide.

Study design and population

In this cohort study, the records of 6289 CAD patients receiving coronary revascularization were obtained in Medical Information Mart for Intensive Care IV (MIMIC-IV). MIMIC-IV builds upon the success of MIMIC-III and incorporates numerous enhancements from 2008 to 2019. MIMIC-IV is a relational database that encompasses authentic hospitalizations of patients admitted to a tertiary academic medical center located in Boston, MA, USA. Each patient’s length of stay, laboratory tests, medication treatment, vital signs and other comprehensive information during their ICU stay were recorded.

A total of 6289 CAD patients undergoing coronary revascularization were identified in MIMIC-IV. Among them, patients with the length of ICU stay less than 24 h were excluded (n = 532). Finally, 5757 participants were included. All patients were divided into the postoperative stroke group (n = 433) and postoperative non-stroke group according whether postoperative stroke occurred.

The present study constructed several prediction models for the risk of stroke in CAD patients who received coronary revascularization based on machine learning methods. The results demonstrated that Catboost model was the optimal model for predicting the risk of stroke in CAD patients who received coronary revascularization. The AUC of Catboost model was 0.831 in the training set, and 0.760 in the testing set, which were higher than the logistic regression model. The findings might provide a novel and quick tool to identify CAD patients receiving coronary revascularization treatments who were at high risk of stroke, and offer timely interventions to prevent the poor prognosis.

Citation: Lin L, Ding L, Fu Z, Zhang L (2024) Machine learning-based models for prediction of the risk of stroke in coronary artery disease patients receiving coronary revascularization. PLoS ONE 19(2): e0296402.

Editor: Aamna Mohammed AlShehhi, Khalifa University, UNITED ARAB EMIRATES

Received: September 4, 2023; Accepted: December 12, 2023; Published: February 8, 2024

Copyright: © 2024 Lin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The datasets generated and/or analyzed during the current study are available in the MIMIC-IV database,

Funding: The author(s) received funding from Dalian Medical Science research project (No. 2112012) for this work.

Competing interests: The authors have declared that no competing interests exist.