Cardiac flow analysis using magnetic resonance imaging.

Abstract

Many types of cardiac abnormality have an implication on blood flow. However, most present-day diagnostic modalities analyse myocardial structures and not the cardiac flow within to detect heart defects in vivo. Currently, various imaging modalities, such as echocardiography, single photon emission computed tomography (SPECT), positron emission tomography (PET), X-ray computed tomography (CT), and cardiac magnetic resonance imaging (CMRI) provide a non-invasive approach for scanning humans with heart abnormalities, and are utilised in the management of cardiac patients. There is a need to develop a visualisation system for analysing flow of blood within the human heart. Motional properties of blood can be measured against normal controls and patients with cardiac abnormalities in order to discover underlying cause of these flow phenomena. This can potentially extend medical knowledge of the defects and their hemodynamic behaviour. We characterise motion patterns of blood in the human heart and analyse the flow properties, by means of tracking, using a series of time dependent magnetic resonance images. An indication of flow vortices can be provided by numerical computation of vorticity values within the defined region of blood flow. The global estimation of parametric motion flow fields over the whole image provides useful information on the presence of vortices within the heart chamber that can be used to assess cardiac functions. In this study, the crucial strategies for this approach are implemented, and the achievable diagnostic results and quality of assessment are investigated. The developmental stages of the framework and system design of each component for cardiac diagnosis are detailed in this thesis. The key objectives of the research and development for this diagnostic system are implemented herein: 1. Realisation of a non-invasive technique to compute flow features within cardiac structures. System evaluation and velocity calibration of the flow tracker are incorporated in the study. Verification of calculated flow in time-resolved cardiac vessels is performed by error analysis using flow fields constructed by velocityencoded magnetic resonance imaging velocimetry. 2. Measurement of cardiac vorticities in heart chambers is performed for investigation of flow phenomena. We examine the time-dependent behaviour of cardiac flow structures in the heart. The variation of flow patterns that are associated with myocardial wall deformations and pressure changes is analysed. 3. Realisation of a statistical framework for examining variations of flow due to myocardial defects in the heart. The quantification of flow will offer the potential to complement diagnostic methods that analyse cardiac defects and evaluate patient condition after surgical intervention. As an alternative to established medical imaging-based diagnostic techniques such as chest X-rays, and pulsed or continuous wave Doppler ultrasound scans for cardiac diagnosis, we develop a magnetic resonance imaging based approach and perform flow quantification to analyse the heart, vis-`a-vis blood movement in chambers based on a measured flow field. This framework offers potential for non-invasive flow visualisation in cardiac structures. We validate this methodology specifically for analysing flow characteristics within a human heart case study. We also demonstrate the potential for non-invasive assessment of cardiac abnormality for a pathological case of the heart.