This document discusses applying a 1D convolutional neural network (CNN) to classify electrocardiogram (ECG) signals and detect cardiac abnormalities. Specifically:
1) A binary classifier using a 1D-CNN is implemented to detect suspect anomalies in ECGs, regardless of the type of cardiac pathology.
2) The model is tested on 21 categories of different cardiac pathologies classified as anomalies.
3) The goal is to have a system that can be implemented on low-cost portable devices to help detect potential cardiac issues from ECG readings.