| X-ray have laid a solid foundation for the invention of X-CT machine. Raising the low-dose CT is not only a necessity of the CT development, but also a trend of medical imaging, radiology development. The low-dose CT images have the features of low PSNR and poor quality, which has a direct effect on identifying the patient's lesions in a medical image for doctors. Therefore, low dose X-CT image quality optimization has become a research focus.X-CT development process, the reconstruction of CT imaging principle and the principle of CT reconstruction are introduced in this paper. In the experiment,we add Gaussian noise to a general grayscale image,and then Wiener filter can be used to denoise the image. The result shows that the Wiener filtering method has a good effect on the images with Gaussian noise. Secondly, the noise contained within the low-dose X-CT images will be analyzed. Based on analysis of the relationship between image quality and radiation dose, we get the characteristic of the noise distribution, which conforms to the Gaussian distribution. On this basis, wavelet theory can be used for multi-resolution analysis of low doses X-CT images. The low-dose X-CT image is decomposed into four sub-images, including one low-frequency image and three high-frequency images. The low-frequency image can be decomposed into another four sub-images. Then denoise all the high-frequency sub-images with the multi-stage Wiener filter. Finally, low-dose X-CT images will be reconstructed by Wavelet Transform. Finally, different doses of X-CT images will be processed in turn, and then different datas with different peaks signal to noise ratio can be acquired.The results of experiment and evaluation show that the method mentioned in this paper has good effects on the low-dose X-CT images such as 10mA,30mA,50mA et images. The method not only improved the peak signal to noise ratio of the image, but also has a significant improvement in visual quality. Therefore, the multi-stage Wiener filter based on multi-resolution analysis designed by this paper has optimal results to the low-dose X-CT images. |