Sinogram restoration followed by filtered backprojection (FBP) is a fast me
thod for obtaining PET images of good quality and quantitative accuracy. In
previous studies, we used detector motion and the method of projection ont
o convex sets for sinogram restoration to improve on FBP reconstruction; bu
t, the performance was observed to degrade rapidly when the noise level of
PET data increased. In this paper, we describe a sinogram restoration metho
d based on Kalman filtering that provides promising results in computer sim
ulations. The potential benefits of this new method includes improved noise
performance and a low computational cost which is comparable to that of co
nventional FBP.