Medical Image Multi-stage Classification Algorithm Based On The Theory Of Symmetric

The development of medical images acquisition and storage technology has led to the rapid growth of the relevant data.In recent years,because of the medical images contains abundant knowledge of images and medical information,medical image data mining technology became a hot issue in medical science and computer science.It can assist physicians to detect and locate pathological changes and determine its benign and malignant.Therefore they are widely used in the clinical diagnosis process.However,there are many factors in the process of the physicians’ diagnosis.The doctors have different knowledge background,even the same medical images could be different judgments.So,using data mining methods to research on medical image classification algorithm to assist doctors in diagnosis the medical images,it improves the efficiency and precision.It has high academic value and practical application.The structures text messages of the patient’s and unstructured information on the medical images provides abundant resources for medical image data mining.The human brain can be divided into two hemispheres with an approximate bilateral symmetry,where most structures in one side have a corresponding counterpart on the other side with similar shape and relative location.The two hemispheres can be distinguished visually by the longitudinal fissure,which is a membrane between the left and right hemispheres filled with cerebro-spinal fluid.If the brain CT images of pathological changes,it will cause the change of symmetry approximation.The presence of the tumor leads to its internal structure exists asymmetry.Based on this medical knowledge guidance,a medical image multi-stage classification(MSC)based on the theory of symmetry is presented in this paper.First of all,weak symmetry and strong symmetry is defined to describe the symmetry from the different granularities.Then the weak symmetry decision algorithm(WSDA)based on the gray histogram intersection was given and the strong symmetry decision algorithm(SSDA)based on the points symmetric was given.WSDA classified medical images into normal and abnormal in coarse granularities.And SSDA classified abnormal medical images into left(right)abnormal in fine granularity.And then the features of the lesion area were extracted with the aid of libsvm realizes the classification of benign and malignant lesions area.Lastly,to simulate the process of the doctor’s diagnosis,this paper proposes a multi-stage classification method.They are normal or abnormal(MSC-1)and located the lesions(MSC-2)and the nature of the lesions(MSC-3).From coarse granularity to finer granularity realized step-by-step methods to simulate the doctor’s diagnosis process.Experimental results show that a multi-stage classification method based on the theory of symmetry can increase the accuracy of the classification and reduce the time of the doctor’s diagnosis compared to other methods.