Adaptive correction of scatter and random events for 3-D backprojected PET data

Abstract

Backprojected images offer compact and accurate storage of three-dimensional (3-D) positron emission tomography (PET) data and so are potentially very suited to high-resolution 3-D PET scanners. However, there are very few methods available for correcting backprojected images for the effects of scatter and random events. Scatter correction methods usually involve experimental measurement of scatter response functions or accurate simulations, and are thus limited by the accuracy of the model used. In contrast, the approach proposed here uniquely adapts a general form of a combined scatter and random response function. This method inherently does not rely upon accurate determination of the scatter and random response functions but does, however, rely upon knowledge of the attenuating object's outline. The technique is able to adapt to different imaging situations and account for detector scatter, and in particular is readily implemented in backprojection space. The method has been implemented for a small-volume HIDAC (II) PET scanner.