A noise suppressing iterative reconstruction technique for positron emission tomoghraphy.

A multiplicative iteration scheme for reconstructing three-dimensional images from shift variant, limited angle data taken by a large area positron camera is proposed. Using simulated data sets the algorithm has been proven to be capable of reconstructing source distributions from a very small number of annihilation events. This suggests the applicability of the technique to the measurement of range distributions of beta⁺-active ions implanted into thick targets and especially to in-vivo dose localization in light ion tumor therapy by means of positron emission tomography.