An improved algorithm for interactive medical image segmentation based on Intelligent scissors

Image segmentation is an important part of computer vision technology, which aims to extract areas of interest for feature acquisition and analysis. The accuracy of medical image segmentation has a direct impact on the decision-making capacity of doctors to diagnose patients' condition. In this paper, the interactive medical image segmentation algorithm based on the intelligent scissors is selected to process the medical image. It can not only combine the subjective experience but also reduce the operation obstacles and speed up the segmentation. The main work of this paper is as follows: 1. Based on the full study of the intelligent scissors algorithm, the principles and methods for the digraph's weight selection in the algorithm construction are clarified, which can improve the accuracy of the algorithm effectively. 2. The IOU factor was introduced to evaluate the image segmentation effect, which can analyze the existing defects from a quantitative perspective. By calculating the similarity between pathological images and normal images, the rationality of image segmentation result analysis is improved. 3. The intelligent scissors algorithm is improved by the modified Canny operator to reduce the generation of pseudo-edges and protect the weaker edges from rejection, which largely improves the accuracy of the segmentation. 4. The research work of the thesis was validated by applying pathological images of patients with novel coronavirus pneumonia and lung cancer. It is proved that the work in this paper has certain theoretical Significance and practical application value.