5D Image Reconstruction for Tomographic Image Sequences

In this paper we propose an image reconstruction procedure which aims to unify gated imaging and dynamic imaging in nuclear cardiac imaging. As in gated imaging the cardiac cycle is divided into a number of gated intervals, but the tracer distribution for each gate is treated as a time- varying signal as in dynamic imaging. By using both dynamic and motion-compensated temporal regularization, our five- dimensional (5D) reconstruction will produce an image sequence that shows three-dimensional (3D) spatial distribution of the tracer, as well as its two-dimensional (2D) temporal changes, i.e. cardiac motion and tracer´s kinetics. To demonstrate the proposed reconstruction method, we simulated gated cardiac single photon emission computed tomography (SPECT) imaging using the NURBS-based cardiac-torso (NCAT) phantom with Tc^99m-Teboroxime as the imaging agent. The results show that our method can reconstruct a meaningful 5D data array and motion- compensated temporal regularization can lead to improved reconstruction than using spatial smoothness alone.