Model-guided Iterative Reconstruction for Limited-angle CT Image

Aiming at the problem of limited-angle projection CT reconstruction, a model-guided iterative reconstruction for limit-angle CT image is proposed in this paper. An initial reconstructed image is firstly gained by statistical iterative algorithm from limited-angle data. The initial reconstructed image is fused with a prior-model image to acquire a fusion image. The fusion image is projected again to acquire the complementary projections, on which are the intermediate results of the reconstruction. This projection, fusion and reconstruction processes repeat until the final satisfactory image reconstructed. Simulation results have shown that the object can be reconstructed integrally by the proposed method and the quality of the final reconstructed images increase significantly, with all characteristics of the object preserved.