Development of simulation tools and enhanced image-guidance for optimized planning and monitoring of high intensity focused ultrasound energy delivery

点击翻译按钮获取中文摘要

Magnetic resonance guided high intensity focused ultrasound (MRgHIFU) is a noninvasive therapy that is approved in Canada for thermal ablation of bone and soft tissue. MRgHIFU transmits ultrasound energy from outside the body and focuses that energy within the body, producing a thermal lesion as small as a rice kernel (8 mm long x 2 mm diameter). This focus is steered electronically and mechanically to cover a large volume, without repositioning the target. Since energy delivery is nonionizing and incisionless, the risk of infection and long-term side effects are minimal. The energy delivery is monitored in soft tissues with MR thermometry (MRT) by measuring signal changes in heated water molecules. Under ideal conditions, namely non-moving soft-tissue targets, MRT monitors temperature in a very fast and repeatable manner with an uncertainty less than 1C. However, there are many conditions, such as the treatment of moving targets (where conventional MRT methods are no longer stable) or the treatment of bone (where the water MRI signal is weak) that can lead to temperature uncertainties greater than 5C. There are also therapies that are not due to thermal effects and do not produce a temperature change. In these instances, MRT is not the ideal monitoring technique; a more direct method to measure the pressure of the sound waves as they interact with the tissue is required. My program investigates methods to improve MRT by separating the MR signal into changes due to temperature and changes due to motion. This process can reduce temperature uncertainties of over 5C, which is unsuitable for therapy, to under 1C when imaging targets in the abdomen. MRT of the abdomen represents a worst-case scenario for monitoring as temperature measurement is corrupted by periodic motion due to the lungs, and aperiodic motion due to the intestines. I am also developing mathematical models for bone HIFU procedures that incorporate simulations of sound wave propagation with a tissue-specific model of heat propagation in bone and soft tissues. These simulations involve segmentation of inner and outer bone surfaces from MRI scans, and an acoustic simulation of the ultrasound wave interaction throughout the bone tissue. This simulation integrates with a tissue-specific model of the temperature reached in the bone, bone marrow, and surrounding tissue. The wave and temperature models are both implemented in a computationally efficient manner that enables full 3D simulation of an MRgHIFU sonication in under one minute. This enables simulations to run within an intraoperative setting without repositioning the target between the planning MRI scans and the energy delivery. Lastly, I am developing methods to image the acoustic radiation force generated by an ultrasound pressure wave using MRI field measurements. This technique, known as ARFI, enables the localization of non-thermal HIFU energy delivery during MRgHIFU procedures.