Blind deblurring reconstruction technique with applications in PET imaging

Abstract

We developed an empirical PET model taking into account system blurring and a blind iterative reconstruction scheme that estimates both the actual image and the point spread function of the system. Reconstruction images of high quality can be acquired by using the proposed reconstruction technique for both synthetic and experimental data. In the synthetic data study, the algorithm reduces image blurring and preserves the edges without introducing extra artifacts. The localized measurement shows that the performance of the reconstruction image improved by up to 100. In experimental data studies, the contrast and quality of reconstruction is substantially improved. The proposed method shows promise in tumor localization and quantification.