Optimization and Model Reduction Methods for Real Time Adaptive Thermal Therapy Planning and Delivery in Ablation Cancer Treatments

This project is supervised by Professor Karen Veroy-Grepl, Professor Martin Grepl and Philips Research Eindhoven, and is financially supported by the European Commission through the Marie Sklodowska-Curie Actions (European Industrial Doctorate, Project Nr. 642445).

Thermal ablation cancer treatments are performed in a minimally invasive percutaneous manner by inserting a probe directly into the tumor – as for example in radiofrequency and laser ablations – or non-invasively by focused wave super-positioning – as in focused ultrasound treatments. The generated heat destroys the cancerous tissue through carbonization. Ablation treatments have many advantages with respect to their applicability to non-resectable tumors and their local minimally invasive nature. However many aspects of the treatment can be further improved by determining the device positioning and power settings and predicting the treatment outcome using simulations of the ablation process. Such personilization of ablation treatments may further reduce the collateral damage to healthy tissue and increasing the success rates of full tumor ablations.

With our work we aim at improving the accuracy and effectivity of ablation treatments by developing reliable and computationally efficient simulations and optimization routines, which can be used not only preoperative in the planning phase, but also during the treatment. This work concentrates on the optimization methods, which can determine the optimal positioning of the ablation device and the power control of the device, as well as the model order reduction of the chosen model, which will enable updates of the simulated outcome in real time.