CT Image Edge Extraction Algorithm And Quality Assessment Analysis

Nowadays, computer technology has deeply infiltrated into medical treatment. The Computerized Tomography (CT) is the product of computer technology and medicine radiation treatment. As CT technology has high resolution capacity, and it is much more accurate than general X-ray scanning, it has been widely used in medical diagnosis and treatment. As the result, there is more and more researches and on CT images processing.It is important and necessary to process CT image and to segment targeting area for better CT image application in clinical diagnosis and medical research. This paper discusses the median filter and mean filter, and analyzes the Sobel gradient operator. Sobel gradient operator is selected to help carry out CT image edge. This paper proposes an improved edge detection algorithm based on Sobel gradient operator. The two-direction gradient in the traditional Sobel gradient operator is extended to four-direction, and the region edge of image is obtained by the best threshold method. The result of image sets testing shows that the improved algorithm has better time efficiency than the traditional algorithm; and the paper verifies that the most accurate way for image edge detection is the optimal threshold method by segmenting the same image with two different thresholds.In order to solve the over segmentation of the traditional watershed segmentation algorithm, this paper proposes an improved watershed transform based on open and close algorithm to segment the CT image. Simulation of traditional watershed algorithm and improved hybrid algorithm is conducted on several image sets separately, and the comparison result shows that the final segmentation area of the proposed algorithm is significantly reduced, and the accuracy is improved.Noise, spatial resolution, density resolution and artifacts are selected to be CT image quality assessment factors based on the analysis of the principle of CT imaging techniques and related literature. At the same time, this paper analyzes the influences of equipment parameters and scanning parameters to assessment factors, and conducts instance analysis to different CT images.