Quantitative Accuracy of Low-Count SPECT Imaging in Phantom and In Vivo Mouse Studies
Figure 3
Plot of % deviation of measured activity values from true activity values with In-111 using the 2 mm pinhole apertures. Each graph represents the full range of activity values measured for all 3 microcentrifuge tubes. The negative bias at low activities introduced by background subtraction is also evident for short frame times in (a) whereas in (c) there is no significant deviation due to the longer frame time with a higher number of counts collected. This bias is eliminated for short and long frame times without any background removal, (b) and (d), respectively. An offset of approximately 8–10% exists due to attenuation differences between measurements of the calibration syringe and the samples.
