Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.741212
Title: Image processing methods for multi-nuclear magnetic resonance imaging of the lungs
Author: Hughes, Paul
ISNI:       0000 0004 7231 7701
Awarding Body: University of Sheffield
Current Institution: University of Sheffield
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis is concerned with image processing methods for multi-nuclear magnetic resonance lung images. Specifically, analysis of hyperpolarised gas (3 He and 129Xe) and dynamic contrast-enhanced proton images is presented. This work is separated into four distinct themes: 1. A semi-automated method of segmenting ventilation-weighted (hyperpolarised gas) and proton (anatomical) images is presented. This method is shown to improve agreement between observers compared to the current method of manual segmentation. 2. The coefficient of variation is investigated as a marker of ventilation heterogeneity for separating healthy volunteers from patients with a range of obstructive disease. The effect of image signal-to-noise ratio, kernel size and kernel dimensionality on the measures are all investigated, along with the effect of the type of sequence used. Furthermore, the difference in metrics acquired from 3 He and 129Xe datasets is assessed. 3. The effect of inflation level on quantitative metrics of lung function and the repeatability of these measures is investigated in healthy volunteers. Additionally, data from a small cohort of patients with asthma and cystic fibrosis, acquired at functional residual capacity plus 1 litre and total lung capacity, are analysed to show the effect of inflation level on metrics in patients. It is shown that lung ventilated volume percentage is highly reproducible in healthy volunteers. 4. An analysis pipeline is developed to investigate ventilation and perfusion matching using a combination of hyperpolarised gas and dynamic contrast-enhanced MRI. This methodology is validated on healthy volunteers and then applied to a small cohort of patients with asthma pre and post administration of a bronchodilator, and a small cohort of patients with chronic obstructive pulmonary disease.
Supervisor: Wild, Jim Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.741212  DOI: Not available
Share: