Patient-specific simulation of internal defibrillation

Abstract

The objective of this study is to investigate the predictive capacity of computational models of electrical defibrillation by comparing the results of patient-specific simulations to clinically determined defibrillation metrics. Finite volume models of the thoracic conductive anatomy and in situ electrodes were constructed for seven patients who received implantable defibrillators. These models were based on segmented X-ray CT images taken shortly after implant. The models were solved for electric field (current density) distributions corresponding to a defibrillation shock. The defibrillation parameters were calculated from these distributions based on critical mass and inexcitability criteria for successful defibrillation. Preliminary results show good agreement between clinical and simulated thresholds for four of the seven patients modeled to date. The defibrillation parameters for the remaining three patients are underestimated. The correspondence between the predicted and measured defibrillation metrics observed in four of the seven patients is encouraging and provides preliminary support to the potential utility of the modeling approach. This approach may allow for patient specific presurgical planning, as well as provide a convenient computational testbed for evaluating new electrode configurations. Although these results are promising additional subjects are needed to further validate the modeling method.