University of Twente Student Theses


Intraoperatieve feedback tijdens MRI-geleide tumorablatie

Linden, S.J.F. van der and Peters, T.R.W.M. and Wagemans, M.E.H.M and Wubbels, M.G.J. (2019) Intraoperatieve feedback tijdens MRI-geleide tumorablatie.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Abstract:Background: Research has shown that magnetic resonance (MR)-guided focal salvage prostate cryoablation is a good alternative for the treatment of local recurrent prostate cancer (PCa). The intraprocedural MR images are used to evaluate the size and shape of the cryoablation area. However, based on these images it is hard to decide if the complete tumor is ablated, since the tumor is barely visible on the intraprocedural images. Currently the progress of the procedure is still analysed by comparing the multi-parametric (mp)MRI with the intraoperative images on sight. In order to achieve a higher success rate, this progress should be followed more accurately. This study aims at finding an objective measurement, which could be determined intraoperatively, based on the ablation and tumor areas. Method: To determine the volumes two Matlab scripts were written, which segmentates the tumor and cryoablation area out of MRI. The segmentation algorithms can be modified by three variables; a threshold for the segmentation, a windowsize for 2D convolution and a rethreshold after the 2D convolution. First, to determine the optima's a range of values was tested in patient cases (n=10), while calculating the DSC. The cryoablation segmentation with the optimal values was validated by calculating the dice similarity coefficient (DSC) in 10 cases. To assess the practical application and validate the tumor segmentation a phantom experiment was conducted. Results: The determined volume of the tumor based on the phantom experiment was 7% higher on average (n=24). The highest average DSC of the cryoablation segmentation was found using a threshold of 0.06 combined with a windowsize of 10, and a rethreshold of 0.3. In this way a DSC of 0.893±0.060 was achieved. Conclusion: An accurate objective measurement for evaluating the progress of the procedure can be achieved using segmentation techniques in Matlab. This proved to be of value in an experimental procedure. However, comparing the location of the tumor on mpMRI with the cryoablation area during the intervention remains a challenge.
Item Type:Essay (Bachelor)
Faculty:TNW: Science and Technology
Subject:44 medicine
Programme:Technical Medicine BSc (50033)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page