Download the full-sized PDF of Highly Accelerated MRI:  Prior Data Assisted Compressed Sensing for Lung Tumour TrackingDownload the full-sized PDF



Permanent link (DOI):


Export to: EndNote  |  Zotero  |  Mendeley


This file is in the following communities:

Graduate Studies and Research, Faculty of


This file is in the following collections:

Theses and Dissertations

Highly Accelerated MRI: Prior Data Assisted Compressed Sensing for Lung Tumour Tracking Open Access


Other title
Linac MRI Hybrid
Dynamic MRI
Compressed Sensing
Image Guided Radiotherapy
Real Time MRI
Prior Information
Lung Tumour Tracking
Linac MR
Type of item
Degree grantor
University of Alberta
Author or creator
Yip, Eugene
Supervisor and department
Fallone, B. Gino (Oncology)
Rathee, Satyapal (Oncology)
Wachowicz, Keith (Oncology)
Examining committee member and department
Steven Beyea (Radiology, Dalhousie University)
Richard Thompson (Biomedical Engineering)
Nick De Zanche (Oncology)
Atiyah Yahya (Oncology)
Department of Oncology
Medical Physics
Date accepted
Graduation date
2017-06:Spring 2017
Doctor of Philosophy
Degree level
Hybrid magnetic resonance imaging (MRI) and radiation therapy devices are capable of imaging in real-time to track intrafractional lung tumour motion during radiotherapy. In the real time tumour tracking treatment scheme, MR images are acquired in real time, then the tumour target is localized using an automatic contouring algorithm, allowing the radiation beam to follow the moving tumour. Highly accelerated magnetic resonance (MR) imaging methods can yield large increases in acquisition speed at the cost of some increase in reconstruction time. If employed effectively, they can potentially reduce system delay time and/or improves imaging spatial resolution, provide flexibility in imaging parameters and allow for the imaging of multiple slices without reducing frame rate. The aim of this thesis is to develop and validate an MR acceleration strategy which can be used to improve real time tumour tracking. First, an in-house tumour auto-contouring software is validated against a gold standard in both phantom and patient data. Once validated, the auto-contouring algorithm is used to validate the MR acceleration strategy developed. The novel MR acceleration strategy, Prior Data Assisted Compressed Sensing (PDACS), combines the advantages of 2D compressed sensing and the view-sharing strategy. Like 2D compressed sensing, it uses L1 regularization to reconstruct images from undersampled k-space. However, the acceleration achievable with 2D-CS is quite limited, as the line-by-line data acquisitions restricts random sampling to only the phase encode direction in k-space. PDACS, improves the reconstruction by adding previously acquired, motion averaged data into the CS reconstruction via an additional penalty term. Our results have shown that this method is superior to 2D-CS, in terms of reduced artifact power and improved tumour tracking metrics. However, PDACS relies on prior data acquired at the beginning of a dynamic imaging sequence, and thus is dependent on the stability of the baseline MR signal. For shorter duration (i.e. 1 minute) dynamic scans, PDACS is shown to be adequate. However, for longer duration scans (3 minutes), PDACS results in a gradual decline in image quality due to drifts in MR signal. An improved implementation, sliding window PDACS, varies the sampling pattern and allows for “prior data” to be continuously refreshed. Using this improved implementation, sliding window PDACS is shown to successfully remove the negative effect of signal drifts from longer scans.
This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for the purpose of private, scholarly or scientific research. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
Citation for previous publication
Yun J, Yip E, Gabos Z, Wachowicz K, Rathee S, Fallone BG. Neural-network based autocontouring algorithm for intrafractional lung-tumor tracking using linac-MR. Medical Physics. 2015;42(5):2296-2310.Yip E, Yun J, Wachowicz K, Gabos Z, Heikal A, Rathee S, Fallone BG. Prior Data Assisted Compressed Sensing: A Novel MR imaging strategy for real time tracking of lung tumors. Medical Physics. 2014;41(8):082301 (12pp.)Yip E, Yun J, Wachowicz K, Gabos Z, Rathee S, Fallone BG. Sliding Window Prior Data Assisted Compressed Sensing for MRI tracking of lung tumours. Medical Physics. 2016; Accepted Author Manuscript. doi:10.1002/mp.12027

File Details

Date Uploaded
Date Modified
Audit Status
Audits have not yet been run on this file.
File format: pdf (PDF/A)
Mime type: application/pdf
File size: 7342527
Last modified: 2017:06:13 12:24:05-06:00
Filename: Yip_Eugene_201701_PhD.pdf
Original checksum: e27adbdc05bc9cfb43b013e40185fc29
Activity of users you follow
User Activity Date