Maximum-likelihood reconstruction based on a modified Poisson distribution to reduce bias in PET

摘要

The AB-ML algorithm [2] and the proposed NEG-ML algorithm can reduce or eliminate the bias in reconstructions from noisy data. AB-ML shifts the Poisson distribution, while NEG-ML modifies it near zero. Both algorithms have a single parameter that defines the balance between MLEM-like and unweighted-least-squares-like behaviour.