P955
Non-invasive mapping: what is the minimal number of electrodes needed to obtain a good spatial resolution of the activation map?

The 12-lead ECG is highly inaccurate for localization of the site of origin of supraventricular and ventricular arrhythmias.  Non-invasive mapping systems (ECGI) based on a high number of electrodes recording the electrical activity on the surface of the torso have already proven good accuracy for mapping different arrhythmic substrates. The aim of this study was to assess what is the minimal number of leads needed to obtain a precise mapping with the ECGI.

This study enrolled 14 patients (9 male, median age 50 (44-58) years) referred to our center for catheter ablation of premature ventricular contractions (PVC). Patients underwent pre-procedural ECGI using the epicardial and endocardial mapping system . This system uses the DICOM images from contrast computed tomography of the heart and up to 28 adhesive electrodes with 8 leads each, adding up to 224 body-surface leads. All patients underwent invasive electroanatomical mapping and ablation with the magnetic navigation system. We analysed the number of recording leads used to construct the non-invasive activation map of the PVCs and the accuracy and the spatial resolution of the map when comparing to the invasive map. We then reprocessed the exam, using progressively less leads until we only left the leads placed in the standard 12 lead ECG positions and evaluated the concordance with the invasive map as well as the spatial resolution. We considered an earliest activation site (EAS) area of 1 cm2 a good spatial resolution and using a ROC curve we calculated the minimal number of leads necessary to obtain a good spatial resolution.

The median number of electrodes used for the initial map was 170 (138-177). Concordance between non-invasive and invasive site of origin occurred in 11 out of 14 patients. The results are presented in the Table. The minimal number of electrodes to have a good spatial resolution was 100. The area of EAS was significantly lower when using more than 100 leads, respectively 0.65 (0.5-1) cm2 versus 3 (1.6-5) cm2, p < 0.001.

The minimal number of leads to achieve a good spatial resolution was high.  Reducing the number of leads resulted in a significant decrease in spatial resolution and a lower concordance rate.

ECGI data

Abstract Figure. Area of EAS according to the N of leads