Validating defibrillation simulation in a human-shaped phantom

Background: We previously developed a computational model to aid clinicians in positioning implantable cardioverter-defibrillators (ICDs), especially in the case of abnormal anatomies that commonly arise in pediatric cases. We have validated the model clinically on the body surface; however, validation within the volume of the heart is required to establish complete confidence in the model and improve its use in clinical settings. Objective: The goal of this study was to use an animal model and thoracic phantom to record the ICD potential field within the heart and on the torso to validate our defibrillation simulation system. Methods: We recorded defibrillator shock potentials from an ICD suspended together with an animal heart in a human-shaped torso tank and compared them with simulated values. We also compared the scaled distribution threshold, an analog to the defibrillation threshold, calculated from the measured and simulated electric fields within the myocardium. Results: ICD potentials recorded on the tank and cardiac surface and within the myocardium agreed well with those predicted by the simulation. A quantitative comparison of the recorded and simulated potentials yielded a mean correlation of 0.94 and a relative error of 19.1%. The simulation can also predict scaled distribution thresholds similar to those calculated from the measured potential fields. Conclusion: We found that our simulation could predict potential fields with high correlation with the measured values within the heart and on the torso surface. These results support the use of this model for the optimization of ICD placements.