Three-dimensional positron emission tomography imaging with 124I and 86Y

Background Impure positron emitters have physical characteristics that degrade image quality compared to conventional positron emitters like F-18. Two impure positron emitters with potentially interesting applications are I-124 and Y-86. The degradation in image quality due to the imperfection of these isotopes is quantified for a human three-dimensional (3-D) positron emission tomography (PET) system. An acquisition protocol to obtain similar image quality as for F-18 imaging is determined by Monte Carlo simulations. Methods The effects of larger positron range, associated singles and the other decay modes on image quality are determined by extensive Monte Carlo simulations of the Allegro scanner. Spatial resolution was evaluated for both isotopes and compared to spatial resolution of F-18. The loss in sensitivity due to triple coincidences was determined as a function of the axial acceptance angle of the PET scanner. The performance of the scanner at low count rates was studied by determining the noise equivalent count (NEC) values for different upper energy thresholds. The image degrading effect of spurious coincidences is taken into account by adding another factor to the NEC calculation. This allowed the contribution of spurious coincidences to be minimized by using a setting for the appropriate energy window. For this optimal energy window the amount of spurious and scattered coincidences was quantified. Simulations of count rate performance were also done to determine the peak NEC and the activity at which the maximum occurred. Results Spatial resolution degradation, compared to F-18, is about 0.5 mm for Y-86 and 1 mm for I-124. Associated singles have a similar effect as scattered coincidences, as they also add a background to the image. The effect, however, is less important than the effect of scatter. The fraction of triple coincidences is quite small for a 3-D PET scanner used for humans as the axial acceptance angle is still moderate. For the Allegro with an energy resolution of 18% the optimal upper energy threshold was determined at 600 keV. For I-124 this leads to 2.5% extra contamination that needs to be added to the scatter fraction. For Y-86 this fraction is about 5.5%. Conclusion 3-D PET images of I-124 and Y-86 have lower spatial resolution. For PET scanners used for humans the difference is not as important as for scanners used for animals. The limited positron decay fraction of both isotopes can be compensated by increasing the imaging time by a factor of 3-5 (same activity). A short coincidence window limits the contamination from other decay modes. Good energy resolution allows setting a selective upper energy threshold to limit the effect of spurious coincidences. With an appropriate setting of the energy window it should be possible to obtain good image quality in a relatively short time because of the high sensitivity of 3-D PET scanners.