Validation of the quantitative gated SPECT wall thickening map, using the MCAT phantom

Myocardial wall thickening (WT) is a useful parameter for the assessment of cardiac function. The Quantitative Gated SPECT (QGS) package from Cedars Sinai produces a polar map of %WT, along with other results. This paper seeks to validate the QGS WT map by comparison to an absolute standard provided by a set of voxel phantoms. The Mathematical Cardiac Torso (MCAT) phantom (University of North Carolina) was used to generate realistic thorax phantoms from which simulated SPECT images could be generated. Phantoms with and without clinically realistic perfusion defects were constructed over the cardiac cycle, each with a different amount of left ventricular WT (15%, 32%, 51% and 87%). Smoothing was performed using direct 3D convolution with a gaussian kernel (FWHM 20mm) to simulate gamma camera imaging characteristics. The resulting images were analysed using QGS on an Odyssey workstation (Marconi Medical) and %WT maps were produced. These maps were assessed using a 20-step quantitative colour scale. A total of 24 samples were extracted from each polar map with 8 samples taken from apical, mid and basal slices The results were then compared to the reference values. The estimates made by QGS for the phantoms with and without a perfusion defect were satisfactory, with the largest difference from the reference value being found in the phantoms with a large %WT value. Thus the QGS package has been shown to estimate WT with acceptable accuracy, even in the presence of a clinically significant perfusion defect. Copyright (2001) The Australian and New Zealand Society of Nuclear Medicine Inc