Acceleration of a model based scatter correction technique for Positron Emission Tomography using high performance computing technique

Positron Emission Tomography (PET) is a widely used and powerful metabolic imaging technique for functional diagnosis of organs. Compton scattering is a physical effect that results in distortions in the reconstructed image. The model based, so-called, Scatter Simulation (SSS) algorithm is an appropriate solution for scatter correction. However, the SSS algorithm is extremely computation intensive. The application of the SSS algorithm in clinical environment requires the application of high performance computing (HPC) techniques. In this work we give a survey about different high performance computing techniques and introduce the selection process of optimal HPC platform for the implementation of the Single Scatter Simulation algorithm.