Accurate model-based high resolution cardiac image reconstruction in dual source CT

Cardiac imaging represents one of the most challenging imaging problems, requiring high spatial and temporal resolutions along with good tissue contrast. One of the newest clinical cardiac CT scanners incorporates two source-detector pairs in order to improve the temporal resolution by two-fold. To achieve the highest spatial resolution, reconstructions using iterative techniques may be desired. Yet the complexity of the dual-source geometry makes accurate system modeling a challenge. In this paper, we present a model-based iterative reconstruction algorithm for the dual-source CT. We demonstrate, using a total variation formulation, the results of our reconstructions. To accelerate the processing and enhance the quality of the result, we also incorporate a simplified detector response function in the forward projector. A segment of heavily-calcified coronary artery is used to demonstrate the spatial and temporal resolution of this approach with the dual-source system.