Positron Emission Tomography (PET) has become a valuable tool with a broad spectrum of clinical applications in nuclear imaging. PET scanners can collect in vivo information from positron radiotracer distributions, which is further recon- structed to a tomographic image with the help of well established analytical or iterative algorithms. In this current work, an innovative PET image reconstruction method from raw data based on a simple mathematical model is presented. The developed technique utilizes the accumulated density distribution in a predefined voxelized volume of interest. This distribution is calculated by intersecting and weighting the two-gamma annihilation line with the specified voxels. In order to test the efficiency of the new algorithm, GEANT4/GATE simulation studies were performed. In these studies, a cylindrical PET scanner was modeled and the photon interaction points are validated on an accurate physical basis. An appropriate cylin- drical phantom with different positron radiotracers was used and the reconstructed results were compared to the original phantom.
Evaluation of lesion detectability in positron emission tomography when using a convergent penalized likelihood image reconstruction method
