A number of heating sources are available for minimally invasive thermal therapy of tumours. The purpose of this work was to compare, theoretically, the heating characteristics of interstitial microwave, laser and ultrasound sources in three tissue sites: breast, brain and liver. Using a numerical method, the heating patterns, temperature profiles and expected volumes of thermal damage were calculated during standard treatment times with the condition that tissue temperatures were not permitted to rise above 100 degrees C (to ensure tissue vaporization did not occur). Ideal spherical and cylindrical applicators (200 microm and 800 microm radii respectively) were modelled for each energy source to demonstrate the relative importance of geometry and energy attenuation in determining heating and thermal damage profiles. The theoretical model included the effects of the collapse of perfusion due to heating. Heating patterns were less dependent on the energy source when small spherical applicators were modelled than for larger cylindrical applicators due to the very rapid geometrical decrease in energy with distance for the spherical applicators. For larger cylindrical applicators, the energy source was of greater importance. In this case, the energy source with the lowest attenuation coefficient was predicted to produce the largest volume of thermally coagulated tissue, in each tissue site.

This study examines the effectiveness of a single, first-order Arrhenius process in accurately modelling the thermally induced changes in the optical properties, particularly the reduced scattering coefficient, mu(s)', and the absorption coefficient, mu(a), of ex vivo rat prostate. Recent work has shown that mu(s)' can increase as much as five-fold due to thermal coagulation, and the observed change in mu(s)' has been modelled well according to a first-order rate process in albumen. Conversely, optical property measurements conducted using pig liver suggest that this change in mu(s)' cannot suitably be described using a single rate parameter. In canine prostate, measurements have indicated that while the absorption coefficient varies with temperature, it does not do so according to first-order kinetics. A double integrating sphere system was used to measure the reflectance and transmittance of light at 810 nm through a thin sample of prostate. Using prostate samples collected from Sprague Dawley rats, optical properties were measured at a constant elevated temperature. Tissue samples were measured over the range 54-83 degrees C. The optical properties of the sample were determined through comparison with reflectance and transmittance values predicted by a Monte Carlo simulation of light propagation in turbid media. A first order Arrhenius model was applied to the observed change in mu(s)' and mu(a) to determine the rate process parameters for thermal coagulation. The measured rate coefficients were Ea = (7.18 +/- 1.74) x 10(4) J mol(-1) and Afreq = 3.14 x 10(8) s(-1) for mu(s)'. It was determined that the change in mu(s)' is well described by a single first-order rate process. Similar analysis performed on the changes in mu(a) due to increased temperatures yielded Ea = (1.01 +/- 0.35) x 10(5) J mol(-1) and Afreq = 8.92 x 10(12) s(-1). The results for mu(a) suggest that the Arrhenius model may be applicable to the changes in absorption.