Imaging performance of a hybrid x-ray computed tomography-fluorescence molecular tomography system using priors.

摘要

PURPOSE: The performance is studied of two newly introduced and previously suggested methods that incorporate priors into inversion schemes associated with data from a recently developed hybrid x-ray computed tomography and fluorescence molecular tomography system, the latter based on CCD camera photon detection. The unique data set studied attains accurately registered data of high spatially sampled photon fields propagating through tissue along 360 degrees projections. METHODS: Approaches that incorporate structural prior information were included in the inverse problem by adding a penalty term to the minimization function utilized for image reconstructions. Results were compared as to their performance with simulated and experimental data from a lung inflammation animal model and against the inversions achieved when not using priors. RESULTS: The importance of using priors over stand-alone inversions is also showcased with high spatial sampling simulated and experimental data. The approach of optimal performance in resolving fluorescent biodistribution in small animals is also discussed. CONCLUSIONS: Inclusion of prior information from x-ray CT data in the reconstruction of the fluorescence biodistribution leads to improved agreement between the reconstruction and validation images for both simulated and experimental data.