Cardio Twin: A Digital Twin of the human heart running on the edge

Abstract

We present the Cardio Twin architecture for Ischemic Heart Disease (IHD) detection designed to run on the edge. We classify non-myocardial and myocardial conditions with a CCN. This CNN generates features from the electrocardiograms and performs the classification task. The database used is "PTB Diagnostic ECG Database" from Physio Bank and it comes from 200 different people. Each patient data sample was partitioned into 2.5 second windows for training. The implemented model achieved 85.77% accuracy and used 4.8 seconds for each sample classification. The results show that technology is ready to fully support demanding processes, such as Digital Twin, on the edge.