Characterization of tumour temperature distributions in hyperthermia based on assumed mathematical forms

Assessing the efficacy of hyperthermia treatments involves three distinct problems: (1) adequately sampling the spatial temperature distribution in a region; (2) defining (a set of) 'descriptors', numerical values which could be used in comparing distinct treatments; (3) testing whether the predictions of prognosis are statistically significant. This paper addresses the first two problems. We use simple assumptions about the tumour geometry and heating pattern to obtain convenient mathematical representations of a temperature distribution, which are then used in defining scalar descriptors such as weighted average temperature TV, and the fraction of tumour volume heated above a given temperature VT/V. Two extreme cases are discussed. In the first, tumour geometry plays the dominant role, and in the second the specific absorption rate (SAR) distribution is assumed to have the greatest influence on the temperature distribution.