• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.17 no.7 s.110
• /
• pp.684-693
• /
• 2006

In this paper, the SA (Specific Absorption) distribution in a human head, exposed to wideband pulse EMF, has been analyzed by taking into account the dispersion characteristics of biological tissues. The dispersive properties of biological tissues are characterized by the 4th Cole-Cole model. Currently, there is no dispersive FDTD algorithm to implement the 4th Cole-Cole model accurately. Thus, in this paper the FDTD methods with the dispersive algorithm for the 1st-order Cole-Cole model and the 3rd-order Debye model were used for SA analysis. The validity of each model has been investigated first, and then the effects of dispersion on SA distribution have been studied.

• Geophysics and Geophysical Exploration
• /
• v.10 no.1
• /
• pp.60-68
• /
• 2007

Conventional analysis of spectral induced polarization (SIP) data consists of measuring impedances over a range of frequencies, followed by spectral analysis to estimate spectral parameters. For the quantitative and accurate estimation of subsurface SIP parameter distribution, however, a sophisticated and stable inversion technique is required. In this study, we have developed a two-step inversion approach to obtain the two-dimensional distribution of SIP parameters. In the first inversion step, all the SIP data measured over a range of frequencies are simultaneously inverted, adopting cross regularisation of model complex resistivities at each frequency. The cross regularisation makes it possible to enhance the noise characteristics of the inversion by imposing a strong assumption, that complex resistivities should show similar characteristics over a range of frequencies. In numerical experiments, we could verify that our inversion approach successfully reduced inversion artefacts. As a second step, we have also developed an inversion algorithm to obtain SIP parameters based on the Cole-Cole model, in which frequency-dependent complex resistivities from the first step are inverted to obtain a two-dimensional distribution of SIP parameters. In numerical tests, the SIP parameter images showed a fairly good match with the exact model, which suggests that SIP imaging can provide a very useful subsurface image to complement resistivity.