A Monte-Carlo study of image reconstruction for PET

The authors have developed a Monte Carlo program to simulate event generation of a positron-emission tomography (PET) scanner. They used it to study design parameters of a scanner and image reconstruction methods. Based on the experimental data, it is concluded that there is no need to decrease the detector size beyond a certain value because no further improvement is achieved. There should be a tradeoff as far as the filtering function is concerned, so that the resolution is as high as possible and the artifacts due to high-frequency noise are minimal