Performance analysis of an improved 3D PET Monte Carlo simulation and scatter correction

We are developing an accelerated Monte Carlo simulation of PET that can be used for scatter correction of 3D PET data. Our Monte Carlo technique accurately accounts for single, multiple, and dual Compton scatter events, attenuation through the patient bed, activity from outside the field of view, and the true energy response of the scanner. We have incorporated innovative sampling techniques that are compatible with our simulation approach, increasing efficiency by a factor of seven. We have reduced the error in the resulting totals output distributions to <1%. The execution time to acquire 10 million scatter coincidence events for a 3D thorax PET scan is only 4 minutes on a 300 MHz Sun dual processor workstation. We demonstrate that when low noise input data is used, images corrected using the Monte Carlo 3D PET scatter correction demonstrate no statistically significant deviation from true activity concentration when accurate input data was used. The speed and accuracy of our simulation program makes it a powerful research tool and a practical scatter correction for 3D PET clinical imaging.