Research On CT Image Reconstruction Algorithm Based On Incomplete Projection Data

Computed Tomography(CT)technology is an important method for clinical detection and radiation therapy in modern medicine.It is shown that patients exposed to ionizing radiation generated from high doses X-ray are at risk of inducing diseases such as cancer,low-dose CT imaging has received widespread attention.On the premise of ensuring that the CT image imaging quality meets the diagnostic requirements,sparse angle scan method is an effective measure to reduce the X-ray dose.However,the projection data obtained by this under-sampling scan is incomplete,which will cause reduction in the quality of the reconstructed CT image and may lead to a misdiagnosis of the lesion location by the doctor.Therefore,how to reduce the radiation dose while reconstructing high-quality CT images without affecting the accuracy of the diagnosis has important for the development of CT imaging technology.In order to improve the quality of sparse angle CT image reconstruction,the CT image reconstruction algorithm based on the incomplete projection data is researched.The main contributions are as follows:(1)As the edges of traditional Non-Local Mean(NLM)constrained CT reconstruction algorithm tends to be over-smooth at the sparse projection angle,an improved algebraic iterative reconstruction algorithm based on adaptive Non-Local Mean constrained(ART-IANLM)is proposed.First,a new similarity window shape is designed,which is composed of the central pixel and its upper,right,and lower left pixels,and is called a clock-like similarity window.Secondly,the adaptive filtering parameters are designed to filter the reconstructed image,which is determinated by the difference between the gray value of the neighborhood pixel and the center pixel in the clock-like window's three directions as well as the number of iterations,It can remove noise and protect edges.Compared with the ART,ART-NLM,and ART-ANLM algorithms,experimental results show that the reconstructed image of ART-IANLM algorithm is not only closer to the real phantom,but also has a smaller reconstruction error,which can protect the edge characteristics of the image better.(2)In order to solve the blurred structural details and over-smoothing effects in sparse representation dictionary learning reconstruction algorithm,a sparse angle CT reconstruction with weighted dictionary learning algorithm based on adaptive Group-Sparsity Regularization(AGSR-SART)is proposed.Firstly,covariance is introduced into Euclidean distance as a new similarity measure,Non-Local image patches are adaptively divided into groups of different sizes as the basic unit of sparse representation.Secondly,the weight factor of the regular constraint terms are designed through the residuals represented by the dictionary,so that the algorithm takes different smoothing effects on different regions of the image during the iterative process,and the sparse reconstructed image is modified according to the difference between the estimated value and the intermediate image to speed up convergence.Finally,the Split Bregman iterative(SBI)algorithm is used to solve the objective function.Compared with the FBP,SART,TV-POCS,and GSR-SART algorithms,experimental results show that the algorithm presented eliminates the effect of excessive smoothing in sparse angle reconstruction,enhances the sparseness and Non-Local self-similarity of the image,and smooths the noise while retaining low contrast information.