The Study And Application Of Medical CT Image Denoising And Enhancement Method

With the development of science and technology, in clinical medical CT images play a more and more important roles, it has become an important method to diagnose the diseases for doctors, especially in the early detection of lung cancer. In recent years, the developments of medical imaging technology have improved the accuracy and reliability of medical diagnosis. Medical CT images are very different from common images, which is the image of the human body. The density of every body’s tissues and organs are not identical, human respiration and heartbeat and the differences in the quality of image equipment, these factors may cause of region of interest(ROI) contrast in CT image difference, fuzzy edge details, doping noise problems, these problems will have great influence to the doctor’s diagnosis. It is very important to study the denoising and enhancement algorithm of the medical CT image. In this paper, we mainly study the CT image denoising and CT image enhancement, and put forward two kin ds of algorithms that are suitable for the denoisingand enhancement of the CT image. The main work of this paper includes the following two aspects:In terms of CT image denoising. In this paper, we first study some traditional image denoising methods, and found that the traditional image denoising algorithm used to dnoise the medical CT image, processed all pixels of the image, and the noise effect is not ideal, the details of the processed medical CT image will become very blurred, and the CT image lost a lot of detail information. According to the general characteristics of medical CT, this paper proposes a medical image denoising algorithm based on Grey Relational Analysis and GM(grey model) prediction model. Firstly, the gray correlation analysis is used to judge whether the pixel is a noise point, and the GM prediction model is used to deal with noise points. The algorithm is able to avoid processing the pixels that are not noise point, and reduce the damage to the details of the CT image. The medical CT images processed by the algorithm with better MSE and PSNR, can retain more details, more closely to ensure the authenticity of the image.In the CT image enhancement, we first apply some traditional image enhancement methods to enhance the CT image, we found that these methods can not be very well enhanced the CT images to achieve the desired results. In the study of the wavelet transform, the wavelet transform can decompose medical CT image with multi resolution and multi levels., to obtain more detail information of high frequency of the medical CT images, will further enhance the high frequency detail information of CT image; and Laplacian pyramid can also decompose the medical CT image with multi-scale decomposition, to get the high frequency detail information of the CT image. On the basis of this, we put forward a CT image enhancement algorithm based on wavelet transform and Laplacian Pyramid according to the characteristics of the medical CT image. First, the original medical image was decomposed by wavelet transform. Then, the high frequency information of medical image was decomposed by Laplacian Pyramid. Finally, the results of wavelet transform and Laplacian decomposition of Pyramid were used to reconstruct the image. The experimental results show that the method has obvious effect to enhance the medical CT image detail information, and the enhanced CT image has a good MSE and PSNR, and better resistance to noise, and the information entropy is basically unchanged. Therefore, this algorithm is more suitable for the enhancement of medical CT image than the traditional algorithm.