Accuracy of non-invasive anatomical lead localization in CRT patients: BSPM vs 12-lead ECG

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Grant agency AZV, Ministry of Health of the Czech Republic Background Non-invasive, inverse ECG (iECG), estimating the cardiac activation from ECG signals, may be useful both to support CRT implant and optimization. However, most iECG strategies require a 32-250 electrode body surface potential mapping (BSPM), limiting their clinical utility. Objective This proof of concept study evaluates the accuracy of our novel PaceView iECG method to localize the LV/RV pacemaker lead to the cardiac anatomy using either a 96-electrode BSPM or the 12-lead ECG. Methods PaceView generates activation sequences in combination with the equivalent double layer source model to simulate the corresponding ECG signals. An initial activation sequence is iteratively optimized to match simulated and measured ECG signals. A BSPM with 96 electrodes and CT were recorded for every patient with implanted CRT. The CT was used to create patient-specific heart/torso model and to localize the ECG electrodes and CRT leads. From a BSPM, 9 signals were selected to obtain the 12-lead ECG. Both BSPM and 12-lead ECG were used to localize the RV/LV lead as the earliest ventricular activation site. Results In this study 14 consecutive patients with dilated cardiomyopathy were enrolled: mean age 58±9 years, 7 females. The localization error for the RV/LV lead was 10.4±5.4 (max 22) / 11.3±5.5 (max 20.4) mm using the 12 lead ECG and 11.7±6.1 (max 26.2) / 13.5±5.8 (max 24.5) mm for the BSPM. The time used for a 3D activation map computation was 0.9-1.8 s. Conclusion Our preliminary results show that the non-invasive lead localization error, using the 12-lead ECG, is small and comparable with a 96-lead BSPM. Therefore PaceView might be clinically beneficial, by actively localizing the LV/RV pacing sites during CRT implant, and optimize AV or VV delays at CRT follow-ups.