Scatter correction in PET tomographic reconstruction algorithms

Abstract

Positron Emission Tomography (PET) allows quantification of cellular metabolism, and it is one of the most powerful techniques in early diagnosis of heart, oncological and neurological diseases. Images are obtained through the administration of a radiopharmaceutical that marks the metabolic process that will be studied. Gamma rays emitted by radiotracer fall upon detectors set around the patient generating events that are stored for image reconstruction. In order to get quantifiable images, different corrections are made to events, including scatter. With the aim of implement correction by scatter, modeling through Single Scatter Simulation (SSS) algorithm is made at first, getting estimated scatter. Afterwards, correction in emission image is made. For correction, estimation made through modeling is incorporated to reconstruction algorithm MLEM (​maximum-likelihood expectation–maximization)​. Preliminary results, obtained in constant activity volumes simulations, have shown that implementation made improves emission image homogeneity.