A physical and clinical evaluation of a fixed angle nuclear tomographic cardiac imaging system

The conventional planar imaging techniques of nuclear medicine may be limited in terms of diagnostic accuracy by contributions to the image from isotope uptake in non-target organs which under or overlie the organ of interest. Similarly contributions from healthy sections of the target organ may mask under or overlying abnormal sections. Various tomographic techniques have been introduced to overcome this problem by producing images of a number of reconstructed planes at different depths through the target organ. This thesis is an evaluation in both physical and clinical terms of one such device (the 7 pinhole collimator) which is used primarily in the imaging of the left ventricular myocardium and blood pool. The 7 pinhole and another fixed angle tomographic system (the Rotating Slant Hole) are compared and evaluated in terms of planar and depth resolutions, reconstructed and acquired sensitivities, uniformities, reconstruction volumes and the effects of scatter. The Rotating Slant Hole is found to have a better physical performance but its low sensitivity and inappropriate reconstruction volume are considered to impose certain limits an its applicability to the clinical situation. An attempt is made to improve the resolution of the 7 pinhole by applying a non-linearity correction to the raw data. The theoretical and experimental accuracy of this technique is established and is found to significantly improve the system resolution with depth. Software generated distortions are introduced into the raw data in order todetermine the effect of various degrees of non-linearity on the resolution. A technique is developed to quantify uptake segmentally in thallium-201 perfusion images. This technique is used to determine theeffects of malrotated and off-centre acquisitions on the production of artefactual defects during tanographic reconstruction. Studies with a left ventricular myocardium phantom reveal that the 7 pinhole is able to detect smaller lesions than the conventional planar technique but both the non-linearity correction and the impedance operator are found in practice not to improve on this. In a group of 26 patients no significant difference was found however in the diagnostic accuracy of the 7 pinhole and planar techniques in the diagnosis of myocardial infarction.The 7 pinhole system is used for blood pool analysis in a dynamic phantom model and in patients . A study of 20 patients revealed the technique to be as accurate as the planar technique in the detection of wall motion abnormalities and more accurate in the identification of dyskinetic segments provided phase analysis was used in conjunction with the reconstructed blood pool images. The model studies reveal the limits of the systems tanographic ability in the case of aneurysmal ventricles.