Abstract
1661
Objectives: To assess the accuracy of CT-based quantitative SPECT for Tc99m using transmission-dependent scatter correction (TDSC) in conjunction with corrections for attenuation (AC), dead time and partial volume effect (PVE) in phantoms and in human studies.
Methods: All studies were performed on a Philips SKYLight gamma camera with in-line single-slice spiral CT (Picker PQ5000) and HERMES workstations. The CT data are converted to an attenuation map which is used to perform TDSC. The scatter corrected projection data are then reconstructed using OSEM with AC and dead time correction. A camera/collimator sensitivity calibration factor is used to transform pixel values to units of radioactivity. The system recovery coefficient has been modelled to allow PVE corrections where necessary. Quantitative evaluation was performed with (a) a cylindrical phantom filled with Tc99m, and (b) an anthropomorphic torso phantom, containing concentrations of 481kBq/ml in the liver region and 270kBq/ml in the inner chamber of the cardiac insert. Further evaluation was performed to assess clinical lung ventilation-subtracted perfusion (V/Q) SPECT/CT studies to the calibrated injected dose of [Tc99m]-MAA.
Results: The reconstruction from phantom (a) yielded total activity values accurate to within 1% of the known value. Several ROIs from phantom (b) produced average concentrations in the liver and cardiac chamber within 2% and 4% of the true values respectively. The analysis of 12 patient lung V/Q studies yielded an average absolute difference between calculated and true activities of 4%, with a maximum error (underestimate) of ~10%.
Conclusions: The above results confirm the accuracy of the developed CT-based quantitative method. The method will now be extended to include other radiopharmaceuticals.
- Society of Nuclear Medicine, Inc.