Brain Image Segmentation And Feature Recognition Algorithm

With the development of medical imaging technology,computer aided diagnosis technology has been widely used in clinical research.The morbidity of brain diseases is increasing,and brain imaging research has been paid great attention.Through MRI image examination,it can not only prevent the injuries caused by surgical intervention,but also help clinicians.It is of great significance to study the feature recognition of brain MRI images,and the segmentation,feature extraction and feature recognition of brain MRI images are essential steps in medical image processing.In this paper,the fuzzy C mean algorithm is used to segment the image,and the features of the brain image are extracted with the gray level co-occurrence matrix.The multiple feature vectors are classified by support vector machine.The shortcomings of the fuzzy C mean clustering and support vector machines are improved to improve the diagnosis rate of the disease.The main contents of this paper are as follows:(1)The research and analysis of fuzzy C mean clustering algorithm(FCM)is sensitive to noise in brain image segmentation,and the initial clustering center is easy to fall into local extremum.In this paper,genetic algorithm and kernel function are simulated to improve and process it.In this paper,the Gauss kernel function is used to map the input samples to the high dimensional space,and the initial clustering is carried out by using the characteristics of the simulated annealing genetic algorithm to converge the global optimal.The algorithm achieves better segmentation results.(2)To study and analyze the clinical information as one of the features,the texture features extracted with grayscale symbiotic matrix have the advantage of multiple characteristic parameters.We extract the texture of the brain MRI image of the patients with mental illness,and then the average value of the acquired energy,entropy,inertia moment,correlation and so on is used as the texture of the brain MRI image.Sign.(3)Support vector machine(SVM)is applied to recognize the clinical feature information and extract the texture features.Since the generalization ability and robustness of single kernel support vector machine(SVM)are limited,the proposed hybrid kernel support vector machine is used to identify the extracted brain texture features,and the generalization ability of the global kernel function is strong by using the polynomial kernel function method.The RBF kernel function has the local kernel function learning ability.As a result,the two kernel functions are combined linearly with a newhybrid kernel.This new hybrid kernel has the advantages of strong generalization ability and good learning ability,which can better identify the texture features of the brain.The algorithms used in this paper have been verified on the MATLAB platform and achieved good results.Therefore,it not only has a good reference value in the brain MRI image processing,but also has important research significance in the prevention and treatment of brain nerve diseases.