2-DIMENSIONAL STATISTICAL-MODEL FOR REGULARIZED BACKPROJECTION IN SPECT

In SPECT, both the noise affecting the data and the discretization of the inverse Radon transform. are responsible for the ill-posed nature of the reconstruction. To constrain the problem, we propose a regulari zed backprojection method (REP) which takes advantage of the relations hips existing between the continuity properties of the projections and those of the reconstructed object. The REP method involves two stages : first, a statistical model (the fixed-effect model) is used to estim ate the noise-free part of the projections. Then, the filtered project ions are reconstructed using a backprojection algorithm (spline filter ed backprojection) which ensures that the reconstructed object belongs to a space consistent with that containing the projections. The metho d is illustrated using analytical simulations, and the REP approach is compared to the conventional filtered backprojection. The effect on t he reconstructed slices of the parameters involved in REP is studied i n terms of spatial resolution, homogeneity in uniform regions and quan tification. It is shown that appropriate combinations of these paramet ers yield a better compromise between homogeneity and spatial resoluti on than conventional FBP, with similar quantification performances.