Optimization of Septa Shadowing in Pet Scanners with Retractable Septa Using Monte Carlo Techniques

Abstract

Scatter and random coincidences are of significant concern in Positron Tomography because they lead to the misrepresentation of the activity distribution in the final image. In this regard, interplane septa have long been used in ring positron tomographs to suppress the acquisition of photons oblique to the scan planes. The septa extend from the face of the detector array to the edge of the field of view. Not only do the septa block photons scattered oblique to the scan planes and singles originating in different planes, but they also result in a decreased count rate sensitivity of the tomograph to true coincidence photons. In this work, the Monte Carlo technique has been employed to study the effect of "septa shadowing" with respect to septa length in order to determine an optimal septa length. Sensitivity parameters have been derived by the use of Noise Equivalent Count (NEC) Rate Analysis, as well as count rate sensitivity analysis in order to balance the counting and imaging performance of the system. In addition, energy spectra, sinogram profiles and scatter fraction results are presented to quantify the effects of septa length on the trues, scatter and randoms contributions of the data collected by the scanner. It is concluded that the highest NEC rate is achieved with positron tomographs with no septa and operating in 3D mode.