Improving PET image spatial resolution by experimental measurement of scanner blurring properties

This paper presents a method for the improvement of transaxial spatial resolution in positron emission tomography (PET) images in 2D acquisition mode. The method is based on the experimental measurement of the system blurring properties which are then inserted into an iterative reconstruction algorithm. In order to maintain acceptable the computation time, an ordered subset algorithm OSEM was used. The proposed algorithm was tested on different phantoms and on oncological patients. It showed good performances in terms of contrast improvement and partial volume recovery without increasing the image noise component. The application of the presented method is particularly important in whole body oncological studies, where there is the need to resolve small lesion which can be located at different distances from the scanner axis and in different blurring conditions.