Acquisition of dielectric properties of tissues in the microwave range

This thesis aims to conduct an investigation into the factors that inﬂuence the measurement and interpretation of the dielectric properties of biological tissues in the microwave frequency rage. Firstly, a novel two-stage genetic algorithm is developed that provides accurate parametric ﬁtting results for both standard tissue property models from the literature and measured dielectric data with lowest average fractional error compared to the exiting optimisation algorithms. Next, the eﬀect of logarithmic and linear frequency spacing of data points on parametric curve ﬁtting is examined and quantitatively assessed using experimental dielectric data.Then, an ex-vivo animal study is conducted to evaluate dielectric heterogeneity within biological tissues and anatomically accurate dielectric models are developed for kidney and heart tissues across a wide frequency range of 500 MHz to 20 GHz. The results of this animal study suggest that treating the kidney and heart as homogeneous organs, and generalising a single measurement to the dielectric properties of the entire organ, may not accurately represent the actual properties. The heterogeneity within organs should be considered in order to achieve a fully accurate dielectric proﬁle. Further, the available dielectric data from the literature is collected and compared in an attempt to quantify the diﬀerence between tissue measurements conducted ex-vivo and in-vivo. The results suggest that the available data in the literature is not suﬃcient to make a deﬁnitive conclusion but suggest a potential variation between ex-vivo and in-vivo dielectric properties of biological tissues. Next, the available dielectric data from the literature and experimentally measured data are analysed and quantiﬁed in order to examine the inter-species consistency of dielectric properties of organs. For some organs, there may be no inter-species variation in dielectric properties; however, that assertion is not true for all species and organs. A constituent analysis of biological tissues is then presented using 20 tissue samples and a strong positive correlation between the dielectric properties of biological tissues and their water content is observed. Lastly, a novel dielectric mixture model is developed to estimate the in-vivo dielectric properties of biological tissues based on the water content of the tissue only. The proposed dielectric mixture model is able to predict the dielectric properties of biological tissue with water content ranging from 66 to 81% over the frequency range of 500 MHz to 20 GHz.

Funder

Irish Research Council for Science, Engineering and Technology

Publisher

NUI Galway

Rights

Attribution-NonCommercial-NoDerivs 3.0 Ireland

cbn

Collections

University of Galway Theses (PhD Theses)