Accuracy of quantification in SPECT myocardial imaging

Single photon emission computed tomography (SPECT) of the myocardium accounts for the vast majority of nuclear cardiology procedures performed worldwide. This is mainly due to its wide availability, and its established and clinically validated diagnostic and prognostic accuracy. It is due to the availability of quantitative software allowing for reliable and reproducible assessments of myocardial perfusion and function. For many SPECT images were obtained using the Filtered Back Projection (FBP) reconstruction algorithm. Recently, due major advances in hardware and software developments 3-D Iterative Reconstruction algorithms became widely available in nuclear cardiology departments. These SPECT reconstruction algorithms offer improved spatial resolution and image contrast as well as accurate implementation of attenuation and scatter corrections. More recently, the hybrid imaging modality SPECT/CT became more widely available radiotracer distribution as well as for computed tomography. In recent years, research into the effect of attenuation and scatter corrections has concluded that combining CT attenuation correction with scatter correcting provides a precise quantification of radiopharmaceutical distribution obtained in SPECT research. Although the method has been proven to attain “clean” images for a variety of diagnostic means, its effects under respiratory cardiac motions using SPECT and high resolution algorithms in 3-D iterative reconstruction imaging is yet to be tested. This study aims to determine the efficacy and quantitative accuracy for 3-D iterative reconstruction algorithms when compared to FBP techniques for cardiac SPECT images. The study will attempt to investigate the issue of respiratory induced cardiac motion and to demonstrate its effect on spatial resolution, image contrast and quantification in nuclear cardiology on SPECT images reconstructed using FBP and Flash (OSEM) 3-D algorithms. We carried out two types of investigations involving experimental measurements carried out on three different phantoms, and Monte Carlo simulations using the GATE package. For the experimental part, we used a Siemens Symbia T16 SPECT/CT scanner to acquire SPECT/CT images using a Tc99m point source, a cardiac insert and a cardiac anthropomorphic phantom; data was acquired, reconstructed and analysed using standard clinical protocols available on Symbia T16 SPECT/CT scanner. We also looked at CT based attenuation and scatter corrections to study their effect on image quality and quantification