Revisiting the singles rate method for modeling accidental coincidences in PET

In high resolution PET, recent efforts have been invested in enhancing scanners sensitivity by reducing the low the energy threshold. In this situation, the standard estimation method SR (Eq. 1) tends to overestimate the randoms rate. In this work we have investigated an alternative method, ST, that models better the problem by taking into account true coincidences between detectors. Since true coincidences are not directly accesible, we have also proposed an iterative method, ST-i, able to improve SR estimations by using only measured values: singles count rates and prompts.