Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.634746
Title: Angiography and dynamic contrast-enhanced magnetic resonance cardiac perfusion series for patient-specific perfusion analysis and visualisation
Author: Zakkaroff, Constantine
Awarding Body: University of Leeds
Current Institution: University of Leeds
Date of Award: 2013
Availability of Full Text:
Access through EThOS:
Abstract:
DCE-MRI perfusion series are an important diagnostic tool for the identification of ischemic heart disease. Ischemia can be diagnosed through visual detection of myocardial regions with reduced blood supply in perfusion series. In theory, perfusion series could be used to deduce the location of the coronary stenosis according to the population-based AHA model of coronary blood supply. The mapping of each of the 17 segments in the AHA model to one of the three coronary arteries is intended for establishing a separate diagnosis for each coronary artery. In practice, coronary anatomy varies from patient to patient which is acknowledged as the main limitation of the AHA model. The work in this thesis is aimed at overcoming the limitation~ of the AHA model by providing patientspecific mappings of coronary supply territories which can be used directly in quantitative perfusion image analysis. Accurate mapping between coronary arteries and perfusion information has critical significance for diagnosis and therapy planning by helping to identify the cases where the patient might benefit from cardiac intervention procedures such as angioplasty, coronary stent implantation, and coronary bypass surgery. Coronary arteries cannot be observed in MR perfusion series due to low spatial resolution of perfusion data. Coronary MR angiography, on the other hand, is designed for examining coronary arteries. Fusion of diagnostic information obtained from 2D perfusion and 3D angiography data is the primary topic of research in this thesis. Correlation of coronary arteries and perfusion defects potentially can be established through registration of angiography and perfusion datasets. Registration challenges in CMR data arise from two types of motion: periodic contraction and relaxation of the heart causing myocardium shape variation, and respiratory motion resulting in the displacement of myocardium. The cardiac cycle phase mismatch between perfusion and angiograp.hy data sets precludes the use of the standard deformable registration methods because of the under-constrained nature of the problem. This thesis presents a solution for non-rigid fusion of perfusion and angiography data, the mediated spatiotemporal registration, which uses dynamic MR cine series spanning the complete cardiac cycle. The proposed method overcomes the under-constrained nature of perfusion-to-angiography registration by separating out the deformable registration problem and solving it through phase-to-phase registration of the cine series. Evaluation of the diagnostic accuracy of quantitative perfusion analysis with patient-specific territories against perfusion analysis based on the AHA model and X-ray angiography demonstrates the value of the mediated spatiotemporal registration in the context of perfusion analysis while highlighting its applicability to other areas of cardiac image analysis.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.634746  DOI: Not available
Share: