Aim. The reconstruction of ventricular activation electroanatomical maps (EAM) is crucial for setting parameters of cardiac resynchronization therapy (CRT) devices. Our previously developed algorithm for non-invasive reconstruction of cardiac activation maps using 12-lead ECG and computer tomography data was applied to an innovative approach for analysing CRT device settings and their impact on interventricular dyssynchrony characteristics.
Methods. Our methodology involved using the reconstructed activation map of a left bundle branch block (LBBB) to derive parameters for computational models which were subsequently used to simulate various modes of biventricular pacing in CRT devices. Using data from 5 patients with previously registered LBBB and implanted CRT we investigated the impact of different device settings and corresponding ECGs on dyssynchrony characteristics to determine the optimal configurations of cardiac stimulation modes for individual patient treatment. To validate the resulting activation maps we compared patient ECGs with different device settings and simulated ECGs using Spearman correlation coefficient between the measured patient ECG and the calculated signal.
Results. Our findings showed that shortening the QRS duration often does not necessarily optimize interventricular dyssynchrony. The optimal delay between the left and right electrodes as determined by the interventricular dyssynchrony index was dependent on the localization of the active pole of LV lead. The mean Spearman correlation coefficient between the patient's ECG during biventricular pacing and the simulated ECG was 0.73 (± 0.13).
Conclusions. This preliminary study assessed the feasibility of using non-invasive mapping based on solving forward ECG tasks to optimize CRT device parameters. Our results confirmed that ECGs computed under different device settings have a high correlation and can be utilized for effective tuning of CRT device parameters.