Open Collections

Abstract

Dynamic single photon emission computed tomography (SPECT) is a relatively new imaging method that uses radioactive tracers and tomographic data acquisition techniques in order to quantify temporal changes in regional radiotracer concentrations within a patient. This is important as the rate of change in tracer concentration within an organ often can be related to the functional ability of that organ. In this work, a new method is presented that is able to determine these kinetic rates while using a conventional single or multiple detector SPECT camera, and more importantly, a single, slow camera rotation in the data collection process (herein this reconstruction method will be referred to as dSPECT). This reconstruction method is based on the fact that a temporal change in the activity concentration at a given location can be represented by a linear inequality constraint over time. Two iterative reconstruction algorithms, constrained least squares (CLS) and dynamic expectation maximization (dEM), have been tested using this approach with a variety of computer simulations and phantom experiments. In simulations involving a slow dynamic change of activity, results indicate that the dSPECT reconstruction procedure typically produces kinetic parameter estimates with a 7% error when using projection data acquired with a single 180° rotation of a triple headed SPECT camera system. This error increases to about 15% for data acquired with a dual head system and further increases to about 50% for single detector acquisitions. When simulated with faster dynamic parameters, errors increased slightly to about 8%, 12% and 55% for acquisitions involving triple, dual and single head systems respectively.