Abstract: Single photon emission computed tomography (SPECT) is a diagnostic functional imaging modality wherein the distribution of a gamma-emitting tracer inside the body is estimated from data acquired from around the patient. Traditional SPECT cameras consist of one or more heads which rotate slowly around the patient, and conventional image reconstruction algorithms assume that the tracer distribution is static during acquisition. In this talk I will discuss reconstructing a time-varying distribution of radiotracer from slow-rotation SPECT data, which presents a highly underdetermined reconstruction problem. Recovering an accurate dynamic image from the acquired projection data requires incorporating constraints, such as temporal regularization, into the reconstruction process.
The presented work builds on the dSPECT approach of Farncombe et al.(1999), which uses simple inequality constraints to restrict the temporal behaviour of the reconstructed image. I will first discuss a modification to the dSPECT algorithm which imposes a stronger constraint on the temporal behaviour of the time activity curve (TAC) in every voxel of the reconstructed image. The modified constraint promotes smooth temporal behaviour by restricting changes in the concavity of each TAC. In the second part of the talk, I will discuss artifacts which occur in dynamic images reconstructed from single slow-rotation data. These artifacts occur due to the fact that only a small number of views of the object are acquired by the camera at any one time. Using realistic 3D phantom simulations, as well as real-life dynamic renal SPECT data, I will demonstrate methods for correcting these artifacts. I will conclude by putting the presented work in context with recent changes in SPECT technology.