Analysis of defibrillation efficacy and investigation of impedance cardiography with finite element models incorporating anisotropic myocardium
Increasing defibrillation efficacy by estimating and lowering the defibrillation threshold (DFT) is important in positioning implantable defibrillating electrodes. We analyzed the relation between the experimentally-measured DFTs and the myocardial voltage gradients (VGs) simulated by the 3D subject-specific isotropic finite element models. Our data show a statistically significant correlation between the DFT and the left ventricular VG, with its septal region being most significant (cc = 0.74). The DFT correlation with the RV and atrial VG, however, is not significant. We then estimated the errors in this correlation due to the isotropic simplification in modeling myocardium by elastically mapping published myocardial fiber data into our animal models, and found that the error in the estimated myocardial VG should have only a small effect on the above estimated correlation. Because the correlation is also affected by the accuracy of the measured resistivities of anisotropic tissues, e.g., myocardium and skeletal muscle, we studied the geometric effects of a four-electrode probe on the measurement of anisotropic resistivities. Our simulation shows that the measured tissue anisotropy ratio is decreased by increasing the electrode size relative to the interelectrode spacing, which is supported by our experimental measurements. We have thus provided an equation for estimating such errors from the electrode geometry.We also studied how the band- and spot-electrode location affects the measured impedance change in impedance cardiography using realistic diastolic and systolic human thorax models. The results show that ventricular contraction, the only factor that increases systolic impedance, has a larger effect than the resulting expansion of major vessels and decrease in lung and aortic blood resistivities. When spot electrodes are placed on the anterior chest wall near the heart, ventricular contraction dominates and the measured impedance change at systole represents 82% of the contribution from ventricular contraction. When using band electrodes, the impedance change is a more balanced combination of the four effects. These results suggest that impedance cardiography can not directly measure stroke volume based on the whole thoracic impedance change, while spot electrodes may be more useful in detecting local physiological activities in the ventricles.
- Bioengineering