Abstract:
Cardiac magnetic resonance imaging (MRI) has a wide range of clinical applications, and it is a fast advancing medical imaging modality. The aim of this thesis is to explore some of the applications of cardiac MRI. Three different applications of cardiac MRI are investigated. Firstly, a feasibility study of a MRI-augmented right heart catheterisation (RHC) was conducted. In this multi-disciplinary study, traditional diagnostic right heart catheterisation was combined successfully with MRI flow measurement in three consecutive patients. The feasibility of this hybrid technique was demonstrated without any technical difficulties or complications. Haemodynamic monitoring and recording equipment were customised to allow the RHC to be conducted in the MRI scanner. Testing of the cardiac catheter, investigation of catheter visualisation, and properties of gadolinium were also explored as part of this study. Secondly, a modified MRI flow sequence was investigated. The modified “three-slice” phase-contrast flow sequence allowed flow to be measured simultaneously in up to three locations, and thus had the additional benefit of reducing errors due to temporal changes in flow. In ten healthy volunteers, the modified sequence was found to be as accurate when compared with the original product flow sequence. However the effect of background errors on the three-slice sequence was noted to be very significant and required correction using a stationary phantom. The three-slice flow sequence with background correction was then applied to thirteen patients with a variety of clinical indications. Lastly, myocardial strain quantification using cardiac MRI was investigated. A new sequence providing automated numerical and graphical strain measurements was compared with the current gold standard of myocardial tagging. The new sequence allowed strain to be quantified from a standard non-tagged steady-state free precession cardiac cine sequence, without an additional acquisition for tagging. Global and regional circumferential and radial strain measurements were compared between the new sequence and the gold standard technique in ten healthy volunteers. Global circumferential strain was clinically comparable with tagging (-16.3% vs. -19.2%). Radial strain by tagging was recognised to be inaccurate and was implausibly low for volunteers, making comparisons difficult.