Brain Image Segmentation Based On KPCM Optimization Algorithm

Objective: The prevention and treatment of brain diseases is one of the important directions in the current medical research field.Among them,magnetic resonance imaging(MRI)is widely used for its advantages of non-invasive and low cost.However,because of the complexity of brain structure,the blur of local edge and the inhomogeneity of regional gray,brain MR image segmentation has been a hot and difficult problem.Methods: Fuzzy c-means clustering algorithm(FCM)is widely used in MR image segmentation because of its fast computing speed,unsupervised and easy to implement.However,the algorithm does not use the spatial information in the image.Therefore,we first choose the kernel fuzzy C-means(KPCM)algorithm as the segmentation algorithm,then use the adaptive median filter to filter out the noise in the image,and then use the combined algorithm of genetic algorithm(GA)and particle swarm to determine the initial parameters,so as to avoid the danger of the algorithm falling into local extremum and improve the efficiency of the algorithm.Results: The optimized algorithm is applied to three groups of different experiments,which are simulated images of different gray levels,simulated human MR images with different noise levels and real human brain MR image.The results of three groups show,compared with the traditional optimized algorithm,the optimization algorithm proposed in this paper has the advantages of noise image,better robustness and higher segmentation accuracy.Conclusion: The optimization algorithm proposed in this paper can segment the image with high quality,not only has good robustness to the noise image,but also can make the algorithm converge quickly,improve the segmentation accuracy of the algorithm and improve the segmentation efficiency of the algorithm,which has a certain potential value for the automatic segmentation of brain tissue.