Accurate Fourth-Order Debye Model for the Head Tissues Across the 0.1–1 GHz Band Using Metaheuristic Genetic Algorithm

Abstract

Over the past couple of years, there has been a growing interest in implantable medical devices (IMDs) for diagnostic, monitoring, and therapeutic applications. Wireless power transfer (WPT) to IMDs enables the removal of batteries for further miniaturization of the overall IMDs footprint. Accurate modeling of the dielectric properties is therefore essential and inevitable for dosimetry studies to investigate the safety considerations of the WPT to IMDs. In this paper, an accurate fourth-order Debye model is derived based on a metaheuristic genetic algorithm to represent the dispersive nature of the head tissues across the $\text{0.1}\text{--}1\,\text{GHz}$ , which includes the popular subbands for IMDs. The 46 head tissues are reduced to 20 tissues based on the similar dielectric properties and categorized into four different groups based on the major organs in the head. The developed Debye model shows a better fitting for the measured data compared with the Cole–Cole model for all the tissue groups. Predominantly, the achieved error function between the measured dielectric properties, and the fitted Debye model is less than 0.07% for tissue groups. In general, a total error reduction of more than 60% is obtained for the developed Debye model compared with the Cole–Cole model for all tissue groups. The proposed Debye model can be incorporated into the available computational methods, such as the finite difference time domain (FDTD) method, to calculate the distribution of the electromagnetic fields inside the brain with more accuracy than the Cole–Cole model.