The Study Of Region Of Interest Filter Back Projection Reconstruction Algorithm Based On The Smooth Window Function

Image reconstruction from projections have been applied in many fields of science. In medical applications, in order to reduce the harm of X-ray on human body, the reconstruction of the region of interest (ROI) has become a focus of research.We construct a smooth window function used to replace the original window function of the filter back projection(FBP) algorithm. Then we reconstruct images for the global region and ROI. Because reconstructing the ROI image from using the projection data through the ROI will produce the constant bias. We make continuation for the projection data of the boundary of ROI along the radial direction as the new projection data. We apply this method to reconstruction,compare the reconstructed images with those reconstructed from the projection data without continuation, and find that the ROI reconstructed image is improved. In addition, we make the continuation reconstruction for the ROI image by using Hamming window function and the new constructed smooth window function. The two images are compared, and we give the numerical experiment by the evaluation functions, then find that the smooth window function is better. Finally, we reconstruct the ROI image by choosing different pixel numbers, the result shows that the greater the number, the smaller the error.