The Alcyone CZT SPECT camera. Evaluation of performance using phantom measurement and Monte Carlo simulations

Abstract: Background Coronary artery disease is one of the most common causes of death. A common cardiac examination is to give a visualization of the myocardial blood flow by injecting a radio pharmaceutical and detect the radiation with a camera. The camera has for decades been using thallium doped sodium iodide (NaI(Tl)) as the detector material. New detector materials have been developed because of the physical limits of the NaI(Tl). The new clinically available material is cadmium-zinc-telluride. One manufacturer has placed 19 detector panels with pinhole collimators sharing a common focal point. This camera is relatively new on the market and not thoroughly investigated. Purpose It has been investigated if the image quality could be improved with a narrower energy window, the sensitivity and spatial resolution within the field of view (FOV), whether the image was affected by different source locations was investigated. Finally the detection limit for heart defects in a physical phantom and in XCAT phantom was found. Material & method A list mode acquisition was performed with a Data Spectrum phantom with a cardiac insert containing a basal defect, which simulated a myocardial infarction. The energy window was changed from ±10% to ±5% and the acquisition time was changed acquire the same statistics. The resolution was determined by fitting a Gaussian curve to a line source measurement. The detector specific sensitivity and the sensitivity in the reconstructed images were measured by using a spot marker at different positions. The positioning effect were investigated by placing the Data Spectrum phantom with a cardiac insert without any defects at different positions, the measurements were also performed on a patient. The defect limit was investigated by simulating defects with different positions and wall thickness in a voxel based, CT-scanned, anthropomorphic phantom and a XCAT phantom using a verified Monte Carlo software. Results It was found that a reduction in energy window did not improve image quality and the acquisition time or activity had to be increased 25% in order to compensate of the loss of counts. The spatial resolution was unchanged over the investigated area. The detector specific sensitivity varied up to a factor 2 with the source position, but the sensitivity in the reconstructed images varied only 4%. A translation of 2 cm of the patient gives rise to positioning effect which are great enough to possibly change the diagnosis. The detection limit for defects in the anthropomorphic phantom was determined to be between 15% and 30% of the wall thickness. For the NaI(Tl) camera the detection limit have been found to be 50%. The detection limit for the XCAT phantom was not found.