Exploratory Data Analysis in Cardiac Surgery: Enhancing Risk Prediction with Machine Learning

Within the rapidly advancing field of health care, machine learning (ML) technologies are pivotal for enhancing clinical decision-making processes, particularly in high-risk areas such as cardiac surgery. This study explores the integration of ML algorithms into the risk assessment of coronary artery bypass grafting (CABG) surgeries. Traditional risk prediction models like EuroSCORE have been shown to overestimate actual risks, negatively impacting both decision-making and patient outcomes. This paper argues for the adoption of ML models that incorporate complex datasets from electronic health records to enhance prediction accuracy. The retrospective study analyzed patient data from the Cardiovascular Surgery Department at Medipol Mega University Hospital, focusing on crucial preoperative variables such as age, ejection fraction, and other factors identified as significant by previous studies and expert consultations. It employed advanced data preprocessing techniques to prepare data for ML processing and address missing data. The study demonstrates that advanced ML models can yield more precise and personalized risk assessments, which may translate into improved patient outcomes in cardiac surgery. This research underscores the critical role of sophisticated risk models in enhancing the accuracy and efficacy of clinical evaluations in cardiac interventions.