Limitations and significance of thermal washout data obtained during microwave and ultrasound hyperthermia

Publisher Website
Google Scholar
Abstract
It is common in clinical hyperthermia to calculate an ''effective blood flow'' by neglecting tissue thermal conduction and fitting thermal washout data to a simple, perfusion-dominated exponential model. We have applied this approach to characterize thermal dissipative mechanisms in patients treated at the Harvard MIT Hyperthermia Center, by analysing thermal washout curves which were obtained during treatment sessions by momentarily interrupting the applied heating. Unfortunately, these measurements of ''effective blood flow'' in patient sessions have given inconsistent results. These inconsistencies arise from uncertainties inherent in the clinical situation: the actual thermal boundary conditions and the spatiotemporal characteristics of the heating field. To quantify these observations a Green''s function solution to the tissue bioheat equation has been derived, to enable temperature fields produced by various heating geometries to be easily calculated. This has been applied to the analysis of temperature decay curves following local energy deposition representative of ultrasound and microwave hyperthermia therapy devices. These results show that thermal washout data are as dependent on patient- and session-specific parameters as on tissue properties and perfusion. For measurements of effective blood flow following ultrasonic heating, errors are dependent on the measurement position within the heated volume, heating geometry, and duration of heating prior to the decay; for microwave heating, results are dependent on the position of the measurement point within the heated field, the frequency, and the surface boundary conditions, whether heated, cooled, or insulated. Thus, any effective tissue property calculated without correctly modelling the heating geometry, boundary conditions and initial conditions will be of a qualitative rather than quantitative nature, and may lead to erroneous and misleading conclusions concerning tissue and tumour response.

This publication has 13 references indexed in Scilit: