Deep learning to predict mortality risk from ECG recordings
The aim of the study is to explore the use of deep-learning convolutional neural networks (CNNs) for the analysis of electrocardiograms (ECGs) to predict long-term mortality risk.
The electrocardiogram (ECG) is a physiological signal that represents the electrical activity of the heart. The ECG is a tool used in clinical medicine to provide information on the physiological and structural condition of the heart. Although the acquisition of the ECG recording is well standardized and reproducible, the reproducibility of human interpretation of the ECG varies significantly according to levels of experience and expertise. Computer-generated interpretations of the ECG have been used for several years; however, these interpretations are based on predefined rules and manual pattern or feature recognition algorithms that do not always capture the complexity and nuances of the ECG signal. More recently, artificial intelligence (AI) in the form of deep-learning convolutional neural networks (CNNs) has been deployed to analyze routine 12-lead ECG recordings. CNN's have been successfully developed to detect left ventricular dysfunction, atrial fibrillation, hypertrophic cardiomyopathy, and an individual's age and sex on the basis of ECG alone. Advanced deep-learning techniques have enabled rapid and precise interpretation of ECG signals, making the ECG a powerful, non-invasive biomarker for diagnostics and cardiac event prediction.
Developing risk models using electrograms can be used to better inform the public of risk and help improve mortality.