Classification Of Mild Cognitive Impairment Based On Cerebral Cortex Thickness

Alzheimer’s disease is a typical neurodegenerative disease,the pathological process is complex,and there is no effective treatment to cure the disease.Therefore, the early diagnosis and prediction of the disease is the main means to prevent the disease.Mild cognitive impairment is generally referred to as the precursor symptoms of Alzheimer’s disease,the clinical manifestations of the disease decline in memory and cognitive ability,easily distracted,and judgment ability problems.Compared with Alzheimer’s disease,Its degree of disease is weak.But if you do not treat the disease in a timely manner,it will change to Alzheimer’s disease.Through the study of mild cognitive impairment,we can effectively prevent Alzheimer’s disease.How to accurately classify patients with mild cognitive impairment from clinical trials has been a hot spot in computational neuroimaging.In this article,we will explore the classification of mild cognitive impairment based on magnetic resonance imaging.Finally,we can provide guidance and theoretical basis for the early prediction of Alzheimer’s disease. This paper mainly includes the research and progress:（1）The collected MRI were classified and processed,in which the normal subjects had NC_SC and NC₂4,while the MCI were divided into MCI_SC and MCI₂4.By comparing the NC group and the MCI group,the thickness of the cerebral cortex of MCI patients was explored.We found that the cortical thickness of MCI patients was significantly thinner than that of normal elderly people.So as to provide a theoretical basis for the classification of cortical thickness.（2）Using the method of pattern recognition,the average thickness of the 78 regions of the cerebral cortex was used as the classification feature,and then the NC group and the MCI group were classified.We compared the effects of three feature selection methods on the classification results.One is SVMRFE algorithm,another one is Relief algorithm,the last is mix method based on SVMRFE and Relief.The classification results show that it is feasible to classify MCI and NC with the cortical thickness.