Iterative reconstruction of tomographic images with Fourier components as model parameters

Iterative tomographic image reconstruction methods have focused on the optimization of some measure in the acquired data space with the image pixel values serving as the model parameters. We have developed an alternative approach in which the model parameters are instead pixel values in the two-dimensional frequency space representation of the model image. The results are at least equivalent to those obtained with the model in image space with the advantage of a considerable decrease in computational cost.< >