Abstract
Background
Noninvasive electrocardiographic imaging (ECGi) using the equivalent double layer (EDL) source model enables both epicardial and endocardial reconstruction of electroanatomical activation patterns during sinus rhythm. The EDL source model has been validated in torso-tank models and animal experiments, however validation in humans using invasive electroanatomical mapping is limited.
Purpose
Validation of EDL based ECGi using invasive electroanatomical mapping.
Methods
Ten patients referred for epicardial and endocardial electroanatomical mapping underwent 67 electrode body surface potential mapping (BSPM), cardiac CT imaging and 3D imaging of electrode positions. Anatomical models of the ventricles, lungs and thorax were created and supplemented with electrode positions. Invasive epicardial (4020±1514 contact points), right ventricle endocardial (724±113 contact points) and left ventricle endocardial (459±117 contact points) local activation timing (LAT) maps were compared to ECGi derived LAT maps.
Results
Included patients (mean age 48±21 years, 80% males) were diagnosed with either arrhythmogenic cardiomyopathy (N=4), dilated cardiomyopathy (N=1), symptomatic premature ventricular complexes (N=3) or myocarditis (N=2). Measured BSMP highly correlated with simulated BSPM (97±2%) with a relative difference of 24% ± 6%. Source models consisted of a mean of 2614±214 nodes with a mean distance between nodes of 8.3±1 mm. Overall, invasive LAT maps and ECGi derived LAT maps showed reasonable correlation for epicardial maps (49% ± 18%) and endocardial maps of the right ventricle (48% ± 27%). Endocardial left ventricular LAT maps showed less correlation (23% ± 28). The absolute difference between invasive and ECGi LAT maps was 16.7±14.2 ms for epicardial maps, 17.2±12.9 ms for right ventricle endocardial maps and 34.0±19.8 ms for left ventricle endocardial maps. Visual comparison showed corresponding areas of breakthrough and regions of latest activation in all cases (Figure). Quantitative comparison of these areas showed a mean absolute difference of 33±7mm or 19±10 ms for epicardial maps, 25±10 mm or 17±13 ms for right ventricle endocardial maps and 39±16 mm or 20±6 ms for left ventricular endocardial maps.
Conclusion
Overall agreement was observed between EDL based ECGi LAT maps and electroanatomical LAT maps. Future research will focus on further quantification of this agreement and improvement of ECGi mapping accuracy.
Funding Acknowledgement
Type of funding source: Foundation. Main funding source(s): Netherlands Cardiovascular Research Initiative, an initiative with support of the Dutch Heart Foundation and UCL Hospitals NIHR Biomedical Research Centre.
B-PO02-188 SYNCAVAND MULTIPOINT PACING IMPROVE LEFT VENTRICLE ACTIVATION AS EVALUATED BY ELECTROCARDIOGRAPHIC IMAGING
