Theory of reconstruction of bipolar ECG signal

Sebastian Wildowicz

supervisor: Teodor Buchner

This thesis presents a minimal model, which allows reconstruction of bipolar ECG from a chosen model of the single myocyte action potential. The example shown below (Fig. 1) confirms the ability of the model to reconstruct the bipolar ECG, as the amplitudes of the QRS complex are correctly reconstructed.

The work concerns the issues of modelling the electrical activity of the heart and the development of techniques for the reconstruction of the signal source state. This research is aimed at developing new methods of signal analysis and numerical modelling, the aim of which is to develop medical diagnostics in the area of cardiology and neuroinformatic. The work focuses on Gaussian parameters and how to fit them to recreate the electrocardiographic (ECG) signals, which provide information about the heart's electrical activity and enable the diagnosis of heart diseases. The algorithm is based on the convolution of Gaussian curves with single ventricular action potential as opposed to what is currently developed as methods of signal reconstruction which focus on the reconstruction of epicardial potentials based on known potentials from the thoracic surface. Maps such as the map of electrical activity or isochrones of the activation front of the myocardial surface are used here.