PhD defence I.M. (Iva) Vilas-Boas Ribeiro

Summary:

Hyperthermia consists of increasing the temperature of malignant tumors to the range of 39°C to 44°C. The quality of hyperthermia treatment depends on reliable temperature monitoring supplemented with equipment and simulations that facilitate precise and adaptive treatment application. Despite the advances in hyperthermia treatment planning (HTP), the clinical introduction of Magnetic Resonance (MR)-compatible devices, and their potential, the procedure to adapt and deliver an effective and personalized treatment is far from optimal. The work presented in this thesis investigated treatment modeling, MR temperature monitoring, and the combination of both to enable a more adaptive and effective hyperthermia treatment. These investigations demonstrated the importance of accurate patient imaging and how tissue properties (perfusion) and characteristics (water content and volume) can affect HTP and its reliability for online usage. Furthermore, it was found that gastrointestinal air motion causes severe susceptibilities artifacts making MR-thermometry unreliable for temperature monitoring. A novel approach was proposed to improve temperature monitoring that used model reduction with Kalman filtering and temperature measurements acquired during treatment. This method exploits HTP knowledge for enhancing MR thermometry to compensate for low SNR and a limited sample rate. Therefore, the proposed approach enabled accurate temperature monitoring even in data corrupted by magnetic susceptibility artifacts caused by motion. This work showed that combining modeling and MR-thermometry may enable a more controlled and tailored MR-guided treatment for each patient. Furthermore, the standard use of MR imaging during hyperthermia would enable gathering treatment and temperature-affected parameters, which may further increase treatment quality control.