Accelerated Quantitative Magnetic Resonance Thermometry using an Undersampled Radial Acquisition and an Iterative Reconstruction
Year: 2015
Type: dissertation
Abstract: This thesis investigated the acceleration of MR thermometry, in the context of prostate heating generated via a transurethral ultrasound applicator, using highly undersampled radial acquisitions with a compressed sensing reconstruction. Three clinically-relevant heating patterns were generated in a phantom using a planar ultrasound transducer. Radial k-space was acquired and retrospectively undersampled by acceleration factors of 4, 8, and 16. Thermometry images were reconstructed using zero-filling and a compressed sensing approach and their accuracy was evaluated. A data-driven method was used for selecting the sparsity transform and the tuning parameters were chosen based on representative heating images. At each acceleration rate the compressed sensing thermometry images showed considerable reduction in artifacts relative to the zero-filled case. The Daubechies and Battle-Lemarie transforms were found to work well. At the maximum acceleration factor, thermometry images could be obtained that were of sufficient accuracy for use in the treatment.
Source: TSpace (University of Toronto)
Author: Patrick S. Leonard
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W2616115991