Development of scatter reconstruction algorithms to 3-dimensional positron emission tomography
MetadataShow full item record
In 3-dimensional positron emission tomography (PET), the proportion of the detected photon coincidences arising from scatter can be very high, up to 60%. A standard approach to scatter correction in 3D PET is to estimate the scattered coincidences and subtract them from the list of recorded events. More recently, novel techniques have been proposed in which the information carried by the scattered photons was extracted and directly used in the image reconstruction step. In this work, the algorithms were extended for use in non-TOF 3D PET systems and were qualitatively and quantitatively evaluated. Despite some promising initial outcomes, standard performance metrics of images reconstructed with each technique, representing the quality of images, was shown to be deteriorated under specified conditions. Further work is required to investigate potential benefits of the proposed algorithms for both ideal and clinical conditions.