A Monte Carlo model for positron emission tomography imaging

Abstract

Patient motion during PET imaging acquisitions can result in distortions and artifacts that make accurate interpretation of the image sets difficult. In an attempt to develop a model-based approach, the feasibility of using Monte Carlo (MC) techniques for developing a probabilistic model for PET imaging is investigated. A two dimensional model and its formulation in the Radon space are presented in the scatter-free, scatter, and random noise situations. The reconstructed results using filtered back projection technique demonstrate the feasibility of the approach for clinical application.