Evaluation of actual 3D temperature distributions using a discrete vasculature (DiVa) thermal model

Abstract

For optimisation of hyperthermia understanding of bioheat transfer is mandatory. When modelling bioheat transfer the thermal effect of blood flow need to be accounted for. Our 3D discrete vasculature model (DiVa) does this by modelling vessels discretely. Missing vasculature is described by an additional local heat sink dependent of a 3D perfusion map. Actual 3D temperature distributions from interstitial hyperthermia treatments of brain tumours and of isolated perfused bovine tongues are compared to the results from the treatment planning system. Both discrete vasculature and heat sink approach are used to estimate the temperature distribution. In vivo data acquisition of anatomy, geometry of the heating implant, angiography and perfusion was done with MRI. The angiographic and perfusion data are not yet optimal for hyperthermia treatment planning. However modelling the thermal decay at the end of the treatment yielded perfusion values that are in agreement with the MR findings. Arterial vasculature, heating implant, location of thermocouples and anatomy of the isolated bovine tongues were reconstructed using a cryo-tome slicing technique. The tongues were 'treated' at various perfusion levels, while the steady state temperature profile along 10 tracks was monitored. In this case, with the large vasculature relatively far from the measured profiles both discrete vessel and heat sink simulations showed quantitative agreement with the measured profiles.