Effect of count rate on contrast in a dual head coincidence camera

The sensitivity of signal to background ratio (S/B) to variation in the random fraction of the data as a function of count rate was estimated in a dual head camera geometry setting. The S/B ratio was found to vary up to 100% in the count rate range providing optimum noise equivalent count rate. The variation was found to be dependent on the object geometry, radioactivity distribution and location of the signal and background regions of interest. These results show that it is very important to maintain the same scanning conditions when comparing a repeat to a baseline scan when a random event correction is not implemented before reconstruction. Furthermore we demonstrated that photo-peak-photo-peak only (p-p) acquisition provides superior contrast compared to p-p and photo-peak-Compton acquisition (p-C), when a random correction is not implemented, in spite of its lower sensitivity