Research participants:
B.W. Raaymakers,
J.J.W. Lagendijk, N. Peters**, W.P.Th.M. Mali**, L.W. Bartels**, S.M. Sprinkhuizen**, C.J.G. Bakker**, ** dept. Radiology
The aim of the project is to develop a MRI-guided thermic ultrasound lance for localized and temperature-controlled heat deposition in breast tissues and to demonstrate the utility/pioneer the application of this device for the (preoperative) demarcation of non-palpable lesions (to facilitate surgical intervention/ resection) and, ultimately, for the completely non-invasive thermal treatment of breast tumours.
For this completely non-invasive ablation of breast tumours an absolute control of the thermal dose distribution must be obtained. Any heterogeneity in the distribution may directly result in under dosage of the tumour and unacceptable tumour re-growth. Especially the presence of blood vessels causes localised cooling and so local under dosage. This is shown in figure 1, where an ultrasound heating strategy is applied on a vascularised model of muscle tissue (a bovine tongue in this case).
Figure 1 The experimental setup of an isolated perfused bovine tongue is shown. This tongue has been reconstructed from cryo-microtome slices and the recontruction of the dicrete vasculature is shown together with the HIFU focus. The spiral trajectory of the focus as proposed by Salomir et al. (2000) is then applied to a bioheat model as well as a discrete vasculature model of the tongue. The lower two rows show the impact of the discrete vasculature (lower row) on the resulting temperature distribution compared with the bioheat (upper row).
The expected treatment strategy, designed to guarantee a uniform thermal dose, is to track the tumour supplying vessels using MRA, occlude them using MRI guided SFUS, check the resulting tumour perfusion using dynamic contrast enhanced (DCE) MRI and finally coagulate the tumour area with a significant margin using MRI guided SFUS. The goal must be a sufficient thermal dose in the tumour area. This requires an ultrasound focus that may freely be steered over the full target zone, utilizing the treatment planning and imaging information.
To investigate this treatment strategy we are using our hyperthermia treatment planning system and DIVA thermal model and are developing dedicated thermal modelling for breast tumours, investigate the exact impact of discrete vasculature and perfusion, investigate which vessel generations need to be occluded, the thermal consequences of this vascular occlusion, the SFUS patterns and the resulting thermal dose distributions. Whether this occlusion is also required to lower the total ultrasound intensity and thus to prevent overheating of the entrance window and to reduce the total treatment time, must be investigated. An example of a angiogram of the breast vasculature is shown in figure 2, together with an impression of a reconstructed 3D vascular network, as required for the thermal simulations.
Figure 2 Maximum intensity projection (MIP) of the transversal gadolinium enhanced T1 images of the breast. Also the MIP of the rasterisation of the reconstructed vessels is shown as an overlay on the MIP of the original MRI images. The lower panel shows the 3D reconstruction of the vasculature together with two orthogonal slices through the original MRI data.