Research On Alzheimer’s Disease Detection Method Based On Feature Relationship

Nowadays,a growing number of elderly people are suffering from Alzheimer’s disease and related diseases.Although there is no good cure for this disease,the sooner the disease is diagnosed,the easier it is to control the disease more effectively through the medicine.Meanwhile,the rapid development of medical imaging technology and machine learning technology has been an huge boost for the rapid development of medical research and clinical trials based on them.Therefore,the Alzheimer’s disease diagnostic method based on machine learning image processing technology has been a research hotspot.Due to feature selection is a key stage of Alzheimer’s disease diagnosis in medical image recognition,and feature extraction techniques are relatively mature,so how to select highly discriminative features or feature combinations from original features becomes particularly important,therefore,we research mainly on the feature selection method in this paper.Firstly,introduce the significance of Alzheimer’s disease diagnosis and the current research worldwide;Secondly,the related knowledge of diagnosing Alzheimer’s disease based machine learning is introduced,including related medical imaging technology,diagnostic process and these technologies included in it;Thirdly,the correlation among features is studied,and then a diagnostic method of Alzheimer’s disease based on it is introduced.Finally,on the basis of the in-depth research of how to use high-order correlation among features for feature selection,this proposed a novel classification framework for diagnosing the Alzheimer’s disease based on hypergraph.The specific research work of this paper is as follows:1)This paper summed up and evaluated the Alzheimer’s disease diagnosis related technologies including imaging technology and so on,and analyzed the typical methods.2)On the basis of the comparative analysis of varieties of feature selection technologies and basing on the characteristics of brain medical images from patients with Alzheimer’s disease or mild cognitive impairment,this proposed a diagnostic method of Alzheimer’s disease based on feature correlations,which measures the original feature correlation by the covariance matrix and combines it with the SVM-RFE method for the optimal feature subset generation.3)On the basis of the in-depth analysis of correlation characteristics in features and related high-order correlation measurement techniques,this proposed a novel classification framework for diagnosing the Alzheimer’s disease based on hypergraph which includes the high-order correlation in feature selection process and can improve classification performance.According to the experimental comparison results,the proposed method can be proved to be an effective diagnostic method of Alzheimer’s disease.To be more specific,the method with feature correlation is better than those without it and the method with high-order correlation is better than those with low-order correlation.