CT image reconstruction on Pi-lines in Parallel-Beam

This paper presents an improved CT (Computed Tomography) image reconstruction algorithm, called PB-BPF (Parallel-Beam Back projection Filtration). It mains that the image is reconstructed under the parallel beam scanning mode. It doesn't need to resample to get the image, and it simplifies the usual BPF (Back projection Filtration) in helical cone-beam CT. At First, The projection data is differentiated by using the Shepp-Logan model. Then, the image is reconstructed by choosing parallel PI beams to back-projection, Hilbert transform and constant compensation. At last, the image quality is assessed by using peak signal-to-noise ratio and structural similarity. Being compared with FBP(Filtered Back Projection), the results indicate that the improved algorithm can reconstruct the image accurately, at the same time, it is more close to human visual system.