Hyperthermia aims at increasing the temperature of malignant tissues to the range of 40-44 C. It is used adjuvantly to radiation therapy in order to enhance tumour control and survival as was recently demonstrated for pelvic tumours by the dutch deep hyperthermia group (published in the Lancet, Van der Zee, 2000).
A major problem in quality assurance of hyperthermia is the quantification of treatments. Both the duration and the level of the temperature elevation contribute to the applied dose of a hyperthermia treatment. A first requirement to quantify this dose is a description of the full 3D temperature distribution.
Measuring by means of invasive thermometry does generally not yield a representative sampling. An alternative for assessing the full 3D distribution is the use of thermal modelling. To model heat transfer in solids only conduction has to be taken into account. In vivo temperature calculations also have to cope with convective heat transport by blood. Cold blood enters the locally heated volume, applies cooling, removes heat and by doing so can severely affect the temperature distribution.
Modelling the thermal impact of blood can be done in several fashions as will be discussed in more detail later on. Ideally all blood vessels are taken into account individually. However in clinical applications not all vessels needed for thermal modelling can be reconstructed due to limited data acquisition. This thesis addresses the problem of thermal modelling with incomplete angiographic data and also the experimental validation of discrete vessel simulations.