Effect of lead filters on the performance of a neuro-PET tomograph operated without septa

The value of using lead filters to improve the quality of PET (positron emission tomography) data has been investigated for a CTI/Siemens 953B brain scanner operated without septa. The filters consisted of sheets of thickness 0.5 mm and 1.0 mm and axial width 18 cm, which were placed around the outer surface of the patient port. The port diameter is 35 cm and the axial FO (field of view) is 10.65 cm. The ratios of dead times (with/without filter) with 1 mCi (20 cm uniform cylinder) in the FOV are 0.92 (0.5 mm) and 0.82 (1 mm) at a lower threshold (LLD) of 380 keV; the values at 150 keV are 0.87 and 0.76. Random rates are reduced by 12% (0.5 mm) and 27% (1 mm) at a LLD of 380 keV and by 20% and 40% at 150 keV. These values reflect the significant reduction of lower energy events observed in the measured energy spectrum. However, these reductions are accompanied by a drop in the true coincidence rate such that noise-equivalent counts are reduced for count rates up to the maximum measured trues rate at lower thresholds of 380 keV and 150 keV. It is concluded that the filters used in this study provide no advantage in the performance of the tomograph