Machine learning to predict survival following heart attacks

Associate Professor Sandhir Prasad, RBWH Foundation

2023 Vanguard Grant - One Year

Years funded: 2024 - 2024

Machine learning / artificial-intelligence based tools promise to revolutionise risk prediction in cardiology due to its ability to integrate the prognostic potential of a diverse range of data sources into a single prognostic tool. Patients with heart attacks remain at significantly increased risk of death following their first event. Identifying high risk patients in this setting is critical so that more intensive preventive measures can be applied. Current risk prediction following heart attacks is simplistic and over-reliant on a single measure of heart function – the ejection fraction or EF, even though research has shown that there are other measures of heart size and function that may actually be better than EF for risk prediction. Our study group has built a comprehensive database of 3464 consecutive patients with heart attacks treated over a decade at our institution comprising clinical, angiographic, echocardiographic and outcomes data that can be used to develop machine learning tools to predict survival following heart attacks.