Abstract

We propose a low-dimensional nonlinear model explaining the ECG dynamics, suitable for data compression and possibly feature detection. Tests on real clinically measured ECG signals confirmed very good performance of the model in terms of modeling errors and compression ratio.

Details

Actions