Image reconstruction of cone-beam differential phase contrast CT using phase line integral retrieval

For the image reconstruction of an X-ray cone-beam differential phase contrast CT, a reconstruction technique based on the derivative property of Fourier transform and the FDK algorithm frame was proposed. By which the X-ray cone-beam differential phase contrast CT and absorption-based CT could share the reconstruction software and hardware. On the basis of the phase line integral filter function, the two dimensional differential contrast projection captured by an area array sensor at each rotation sampling view angle was filtered row by row, and the line integral cone-beam projection dataset of phase function for an object was obtained. Then three dimensional phase contrast CT images were reconstructed by applying the classical absorption-based FDK algorithm to the retrieved dataset. With the well-known three dimensional Shepp-Logan phantom, a numerical simulation was performed to validate the proposed method. Finally, the experimental reconstruction was verified by the dataset of a plastic sphere sample and compared with the current methods. Experimental results indicate that normalized root square error of the new method is less than 0.2586 and the reconstruction accuracy is improved by 88%. It satisfies the high-accuracy reconstruction requirements of X-ray cone-beam differential phase contrast CT.