Lung CT Image Segmentation Algorithm Based On Modified 2D-Otsu Optimized By PSO

With the aggravation of atmospheric pollution and the destruction of the environment,lung cancer patients have increased dramatically.Lung cancer has become the most harmful malignant tumor to human health,and the treatment of lung cancer is imminent.At present,the main method of lung tumor examination is to use CT scans.CT images mainly reflect the information of the pathological region.Doctors diagnose and treat patients based on the information of the region.In the study of Computer-Aided Diagnosis(CAD)lung tumors,the segmentation of lung CT images is the first step,and the results of segmentation will have a direct impact on subsequent image processing.Therefore,it is of great significance to improve the efficiency and accuracy of CT image segmentation in lung cancer diagnosis and treatment.In this paper,the two-dimensional Otsu algorithm is used to segment the lung CT image.The traditional two-dimensional Otsu algorithm is used to improve the two-dimensional Otsu algorithm and the particle swarm algorithm is used to optimize the two-dimensional Otsu algorithm,and proposes Lung CT image segmentation algorithm based on modified 2D-Otsu optimized by PSO.Firstly,the imaging principle of CT image and the characteristics of CT image of lungs are introduced,and the principle of one-dimensional Otsu and twodimensional Otsu algorithm is introduced in detail.In view of the disadvantages of large computation and errors in background and target division of traditional twodimensional Otsu algorithm,this paper defines the region of the two-dimensional grayscale histogram of the average of pixel grayscale and neighborhood grayscale,and limits it to the diagonal range from 0 region to 1 region.In order to reduce the search range of two-dimensional threshold and reduce the amount of computation.The experimental results show that the improved two-dimensional Otsu algorithm reduces the threshold segmentation time and improves the efficiency of threshold segmentation.Secondly,Because the exhaustive method used in the calculation of the two-dimensional Otsu algorithm takes a lot of time,and the particle swarm optimization algorithm has a strong ability to find better solutions,this paper selects the particle swarm optimization algorithm to perform threshold search on the two-dimensional Otsu algorithm to reduce the second.Dimension Otsu algorithm operation time.Aiming at the problems of low precision of function optimization,easy to fall into local optimum and slow convergence in the late stage of search in traditional particle swarm optimization algorithm,this paper adopts linearly decreasing inertia weight coefficient and dynamic acceleration coefficient to improve the global and local search ability of particles.To prevent the algorithm from falling into local optimum and improve the convergence speed of the particle swarm algorithm.The experimental results show that the convergence speed and accuracy of the improved particle swarm optimization algorithm are greatly improved.Finally,Using Lung CT image segmentation algorithm based on modified 2DOtsu optimized by PSO proposed in this paper and the other two algorithms to simulate five different lung CT images respectively.The experimental results show that Lung CT image segmentation algorithm based on modified 2D-Otsu optimized by PSO not only improves the segmentation efficiency of lung CT images,but also ensures the accuracy of threshold segmentation.In the case where the CT value of lung tissue is significantly different from other human tissues,the best threshold can be obtained by using this algorithm to perform the initial segmentation of lung CT images,which reduces the impact of threshold selection on segmentation.Then,the external background interference is removed by the hole filling operation.Then,through the boundary tracking method,the boundary of the lung area can be obtained and the lung parenchyma can be extracted.Finally,the pulmonary lesions were repaired by mathematical morphology.Through these operations,the lung parenchyma can be quickly and accurately segmented from the lung CT image,which can be used for computer-aided diagnosis of the lung region.