An investigation into the limitations of myocardial perfusion imaging

Myocardial Perfusion Imaging (MPI) plays a very important role in the management of patients with suspected Coronary Artery Disease and its use has grown despite the shortcomings of the technique. Significant progress has been made in identifying the causes of these shortcomings and many solutions been suggested in the literature but the clinical sensitivity and specificity of the technique is still well below optimum. Monte Carlo Simulation is a very useful tool in identifying and guiding the understanding of the existing problems in MPI and this present study utilised this method to establish the basis of the simulations to be used and the way to analyse the results so that many of the causes of the attenuation defects, when using MPI, could be identified. This was achieved by investigating the effect that the different anatomical parts of the thorax have on the attenuation defects caused. A further aspect investigated was the impact that self-absorption in the heart has on these defects. The variability of these defects were further investigated by altering the position and orientation of the heart itself within the thorax and determining the effect it has on the attenuation defects caused. Results indicate that the attenuation caused is a very complicated process, that the self-absorption of the heart plays an extremely important role and the impact of the different positions and orientation of the heart inside the thorax are also significant. The distortion caused on the images by these factors was demonstrated by the intensity losses in the basal part and an over-estimation in the apical parts, which were clearly observable on the final clinical images, with the potential to affect clinical interpretation. Attenuation correction procedures using transmission sources, have been available for some time, but have not been adopted widely, amidst concern that they introduce additional artefacts. This study determined the effectiveness of these methods by establishing the level of correction obtained and whether additional artefacts were introduced. This included the effectiveness of the compensation achieved with the use of the latest commercially available comprehensive correction techniques. The technique investigated was “Flash3D" from Siemens providing transmission based attenuation correction, depth-dependent resolution recovery and scatter correction. The comparison between the defects and intensity losses predicted by the Monte Carlo Simulations and the corrections provided by this commercial correction technique revealed that solution is compensating almost entirely for these problems and therefore do provide substantial progress in overcoming the limitations of MPI. As a result of the improvements gained from applying these commercially available techniques and the accuracy established in this study for the mentioned technique it is strongly recommended that these new techniques be embraced by the wider Nuclear Medicine community so that the limitations in MPI can be reduced in clinical environment. Non-withstanding the above gains made there remains room for improvement by overcoming the of use transmission attenuation correction techniques by replacing them with emission based techniques. In this study two new related emission based attenuation correction techniques have been suggested and investigated and provide a promising prospect of overcoming these limitations.