Hybrid reconstruction method for exterior CT

In this work, a hybrid reconstruction method is proposed to solve the problem of exterior CT reconstruction. A 2D filtering with data extrapolation method is used to overcome the data truncation problem in exterior CT. The result of 2D filtering is processed by a dual-dictionary learning method to reduce the limited angle artifacts and noise in the exterior CT reconstruction. Experimental studies with simulated projections were performed to evaluate the proposed algorithm. The results reconstructed using the proposed method shows improved image quality.