Few-view tomography using roughness-penalized nonparametric regression andperiodic spline interpolation

The ability to reconstruct high-quality tomographic images from a smaller n umber of projections than is usually used could reduce imaging time for man y nuclear-medicine studies. This would particularly benefit studies such as cardiac SPECT where patient motion during long acquisitions can lead to mo tion artifacts in the reconstructed images. To this end, we have investigat ed sinogram pre-processing techniques designed to enable filtered backproje ction (FBP) to produce high-quality reconstructions from a small number of views. Each projection is first smoothed by performing roughness-penalized nonparametric regression using a generalized linear model that explicitly a ccounts for the Poisson statistics of the data. The resulting fit curves ar e natural cubic splines. After smoothing, additional angular views are gene rated using periodic spline interpolation, and images are reconstructed usi ng FBP. The algorithm was tested on data from SPECT studies of a cardiac ph antom placed at various radial offsets to enable examination of the algorit hm's dependence on the radial extent of the object being imaged.