Quanitifying myocardial blood flow using dynamic contrast enhanced cardiac magnetic resonance imaging

The assessment of myocardial perfusion using dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) is a powerful tool for diagnosing myocardial ischaemia due to coronary heart disease, which affects nearly 2.7 million people in the UK and for which there is an effective treatment. Although visual analysis of DCE-MRI data performs well diagnostically, a quantitative estimate of myocardial blood flow (MBF) makes the diagnosis objective and could increase diagnostic performance. Obtaining MBF estimates from DCE-MRI data is a multi-step process requiring:

- the localisation of the myocardium and arterial input function (AIF) to generate signal intensity vs. time curves;
- the conversion of signal intensity data to contrast agent concentration values;
- the application of a perfusion model to generate a quantitative MBF estimate;
- the interpretation of MBF estimates to make a diagnostic assessment of myocardial ischaemia.

There are a range of approaches for solving each of these problems. The aim of the work presented in this thesis has been to provide clinically relevant evidence for choosing between these approaches. Myocardial localisation contour error tolerance levels are suggested based on simulations using a volunteer dataset. A non-linear signal intensity to contrast agent concentration conversion method is presented and tested using simulations and phantom data. An investigation into the best way to interpret quantitative MBF estimates is then presented. Finally a comparison of four, widely applied, perfusion models is conducted.

Where possible, methods have been compared on a sizeable patient dataset in terms of diagnostic performance rather than MBF estimate accuracy. This provides evidence suitable for informing clinical decisions on the best methodology for quantitative perfusion. Such evidence could contribute to a standard methodology for quantitative cardiac MR perfusion. This is necessary for large clinical trials, which are essential before quantitative MBF estimates can be accepted into routine clinical practice.