Model-Based Relevance of Measuring Electrodes for the Inverse Solution with a Single Dipole

Abstract

Individual ECG electrodes of a multi-lead measuring system can have a variable impact on the solution of the inverse problem. In this study, we investigated the model-based relevance of individual ECG electrodes to identify the position of the stimulation electrode using the inverse solution with a single dipole as an equivalent electrical heart generator. We used four torso ECG mapping datasets from the EDGAR database recorded during ventricular stimulation in three animal torso-tank experiments and one human measurement. The relevance of electrodes, expressed as their weighted contributions to the inverse solution, was determined by the singular value decomposition of a transfer matrix calculated for the given position of the stimulation electrode. The results showed that gradual omission of the electrodes with the highest weighted contributions to the inverse solution worsens the localization. However, missing a small number of such electrodes has little or no effect on the localization. One dataset was more robust to the gradual omission of electrodes with the highest contributions, and the localization significantly deteriorated only after skipping 92% of electrodes. Further study showed that using only several electrodes with the highest weighted contributions to the inverse solution leads to the same or even better localization results than using all electrodes.