Adaptive model-predictive control and 3d acoustic radiation force imaging for the improvement of magnetic resonance-guided focused ultrasound therapies

Update item information
Publication Type dissertation
School or College College of Engineering
Department School of Computing
Author De bever, Joshua Thomas
Title Adaptive model-predictive control and 3d acoustic radiation force imaging for the improvement of magnetic resonance-guided focused ultrasound therapies
Date 2015
Description Focused ultrasound (FUS) is a promising noninvasive and radiation-free cancer therapy that selectively delivers high-intensity acoustic energy to a small target volume. This dissertation presents original research that improves the speed, safety, and efficacy of FUS therapies under magnetic resonance imaging (MRI) guidance. First, a new adaptive model-predictive controller is presented that leverages the ability of MRI to measure temperature inside the patient at near real-time speeds. The controller uses MR temperature feedback to dynamically derive and update a patient-specific thermal model, and optimizes the treatment based on the model's predictions. Treatment safety is a key element of the controller's design, and it can actively protect healthy tissue from unwanted damage. In vivo and simulation studies indicate the controller can safeguard healthy tissue and accelerate treatments by as much as 50%. Significant tradeoffs exist between treatment speed, and safety, which makes a real-time controller absolutely necessary for carrying out efficient, effective, and safe treatments while also highlighting the importance of continued research into optimal treatment planning. Next, two new methods for performing 3D MR acoustic radiation force imaging (MR-ARFI) are presented. Both techniques measure the tissue displacement induced by short bursts of focused ultrasound, and provide a safe way to visualize the ultrasound beam's location. In some scenarios, ARFI is a necessity for proper targeting since traditional MR thermometry cannot measure temperature in fat. The first technique for performing 3D ARFI introduces a novel unbalanced bipolar motion encoding gradient. The results demonstrate that this technique is safe, and that 3D displacement maps can be attained time-efficiently even in organs that contain fat, such as breast. The second technique measures 3D ARFI simultaneously with temperature monitoring. This method uses a multi-contrast gradient recalled echo sequence which makes multiple readings of the data without increasing scan time. This improves the signal to noise ratio and makes it possible to separate the effects of tissue heating vs displacement. Both of the 3D MR-ARFI techniques complement the presented controllersince proper positioning of the focal spot is critical to achieving fast and safe treatments.
Type Text
Publisher University of Utah
Subject Acoustic Radiation Force Imaging; Focused Ultrasound; Model-Predictive Control; MRI
Dissertation Name Doctor of Philosophy
Language eng
Rights Management Copyright © Joshua Thomas de bever 2015
Format Medium application/pdf
Format Extent 27,587 bytes
Identifier etd3/id/3826
ARK ark:/87278/s6mh0xt6
Setname ir_etd
Date Created 2016-06-06
Date Modified 2017-06-29
ID 197377
Reference URL