The Research On Low-dose CT Image Post-processing Algorithm With Improved Non-local Means

X-ray(computed tomography,CT)technology has developed since the last century and plays an important role in many fields of society.For example,industrial testing and medical diagnosis rely on CT technology.However,in medical diagnosis,excessive X-ray radiation will cause serious physiological damage to the tested person,so(low-dose CT,LDCT)technology has become a new technology to replace the traditional high-dose CT technology.By reducing the CT radiation dose entering the patient's body,it can effectively reduce the unnecessary damage to the patient.However,when the radiation dose is reduced,the projection data will inevitably have the disadvantage of a large amount of noise and artifacts,resulting in the image distortion after low-dose CT reconstruction.In order to reduce the noise of the reconstructed image and suppress the artifacts of the reconstructed image.Based on the(non local means,NLM)denoising algorithm,this paper makes full use of the inherent characteristics of low-dose CT reconstructed images by enhancing the original NLM algorithm.The relevant work contents are as follows:1.Firstly,the relevant theoretical basis of NLM noise reduction algorithm is introduced in detail,then NLM noise reduction algorithm and some basic noise reduction algorithms are applied to the brain phantom simulation experiment of low-dose CT,and finally the experimental results are analyzed.The final results show that NLM algorithm has the ability of noise reduction in low-dose CT image post-processing,which reflects that NLM noise reduction algorithm can play its due role in low-dose CT image post-processing.However,any algorithm is not perfect.The classical NLM algorithm has some defects and needs further improvement.2.When NLM denoising algorithm is applied to image denoising,it only considers the structural similarity of local content of the image.When low-dose CT images contain rich texture content,there will be problems such as excessive smoothing at the flat part of the image structure and loss of details at the edge of the image.As a useful measurement tool,(intuitive fuzzy divergence,IFD)is combined with the similarity measurement between image blocks of nonlocal mean denoising algorithm,which effectively improves the problems of classical NLM algorithm.The simulated phantom image and the actual low-dose CT image were studied as experimental objects respectively.The experimental results show that compared with other noise reduction algorithms,the adaptive(intuitive fuzzy divergence non local means,IFD-NLM)algorithm based on intuitive fuzzy divergence has a good effect in the post-processing of low-dose CT images.