Abstract:
Microwave thermography obtains information about the subcutaneous body temperature by a spectral measurement of the intensity of the natural thermally generated radiation emitted by the body tissues. At lower microwave frequencies the thermal radiation can penetrate through biological tissue for significant distances. The microwave thermal radiation from inside the body can be detected and measured non-invasively at the skin surface by the microwave thermography technique, which uses a radiometer to measure the radiation which is received from an antenna on the skin. In the microwave region the radiative power received from a volume of material has a dependence on viewed tissue temperature T(r) of the form, where k is the Boltzmann's constant, B the measurement bandwidth, c(r) is the relative contribution from a volume element dv (the antenna weighting function). The weighting function, c(r), depends on the structure and the dielectric properties of the tissue being viewed, the measurement frequency and the characteristics of the antenna. In any practical radiometer system the body microwave thermal signal has to be measured along with a similar noise signal generated in the radiometer circuits. The work described in this thesis is intended to lead to improvement in the performance of microwave thermography equipment through investigations of antenna weighting functions and radiometer circuit noise sources. All work has been carried out at 3.2 GHz, the central operating frequency of the existing Glasgow developed microwave thermography system. The effects of input circuit losses on the operation of the form of Dicke radiometer used for the Glasgow equipment have been investigated using a computational model and compared with measurements made on test circuits. Very good agreement has been obtained for modelled and measured behaviour. The losses contributed by the microstrip circuit structure, that must be used in the radiometer at 3.2 GHz, have been investigated in detail. Microwave correlation radiometry, by "add and square" method, has been applied to the received signals from a crossed-pair antenna arrangement, the antennas being arranged to view a common region at a certain depth. The antenna response has been investigated using a noise source and by the nonresonant perturbation technique. The received pattern formed by the product of the individual antenna patterns gives a maximum depth in phantom dielectric material. The depth can be adjusted by changing the spacing of the antennas and the phase in an antenna path. However, the pattern is modulated by a set of positive and negative interference fringes so that the complete receive pattern has a complicated form. On uniform temperature distributions the total radiometric signal is zero with the positive and negative contributions cancelling each other out. The fringe modulation can be removed by placing the antennas close enough together, The pattern is then simple and gives a modest maximum response at a known depth in a known material. The radiometer system remains sensitive to the temperature gradients only and the wide range of dielectric properties and tissue structures in the region being investigated usually makes the system response difficult to interpret. For crossed-pair antennas in phase the effective penetration depth in high-and medium-water content tissues is about 2.5 cm at a frequency of 3.2 GHz. The field pattern observed was of the form expected from the measurements of the individual antenna behaviour with the appropriate interference pattern superimposed. The nonresonant perturbation technique has been developed and applied to assist the development of the medical application of both microwave thermographic temperature measurement and microwave hyperthermia induction. These techniques require the electromagnetic field patterns of the special antennas used to be known. These antennas are often formed by short lengths of rectangular or cylindrical waveguide loaded with a low-loss dielectric material to achieve good coupling to body tissues. The high microwave attenuation in biological materials requires the field configurations to be measured close to the antenna aperture in the near-field wave. The nonresonant perturbation is a simple technique which can be used to measure electromagnetic fields in lossy material close to the antenna. It has been applied here to measure accurately the antenna weighting function and the effective penetration depth in tissue simulating dielectric phantom materials. (Abstract shortened by ProQuest.).