Development and validation of a contact force-independent index for predicting pulsed field ablation lesion depth
B Ren, R Zeng, W Li, X Hu, F Li, A Gong, Y Cao, Y Huang, X ZhuAbstract
Background
The lesion depth prediction index has been demonstrated to optimize radiofrequency ablation strategies. Pulsed field ablation (PFA) similarly requires real-time prediction of lesion depth to balance ablation dosage. Currently, most PFA prediction indices rely on contact force. However, contact force-based lesion depth prediction is difficult to integrate into multi-electrode catheters with complex shapes, which represent the current mainstream design for PFA catheters.
Purpose
This study developed and validated a contact force-independent PFA index and assessed its reliability.
Methods
A 3D electric field model was constructed. An impedance-based tissue-proximity index served as a substitute for contact force to evaluate electrode–tissue contact. The predictive model incorporated the tissue-proximity index, number of pulse applications, and voltage. We performed the procedure using a 7-electrode circular PFA system in animal experiments. In six pigs, we injected 250 mL of 1% triphenyl tetrazolium chloride 2 hours post-ablation for staining. Following euthanasia, we excised the hearts, measured lesion depth, and submitted samples for pathological examination. Prediction accuracy was evaluated using the R-value, root mean square error (RMSE), and mean absolute error (MAE).
Results
Computer modeling incorporated real-time impedance variation, voltage, and number of applications to initially establish the algorithm. In the acute phase of animal experiments, both the number of applications (R=0.474, p<0.001) and the tissue proximity indication (R=0.516, p<0.001) were positively correlated with lesion depth. The final integrated model achieved 84.0% accuracy within 1.0 mm (R=0.74, RMSE=0.68 mm, MAE=0.56 mm) (Panel A). Histological and pathological examinations confirmed these lesions (Panel B–C).
Conclusion
The DPI effectively predicts lesion depth without requiring contact force. This provides proof of feasibility for impedance-based prediction of PFA lesion depth and offers valuable inspiration for the future development of predictive algorithms.Panel