Time-varying reconstruction in single photon emission computed tomography

We propose a new method for reconstruction of images in single photon emission computed tomography (SPECT), when the activity distribution of the object is time-varying. The activity evolution is modeled with the first-order Markov model, and linear observation model is used to characterize the measurement system. The state-space representation of the measurement sequence reduces to an ill-conditioned state estimation problem, which is solved recursively using the Kalman filter and smoother algorithms. The method is evaluated using simulations.