The Brain Complexity Of Aging And Alzheimer’s Patients Based On Permutation Entropy

Brain aging is a complex and inevitable biological process.The study of brain maturation and aging growth curve can explore the mechanism of brain aging,and contribute to the early intervention of senile cognitive diseases.Alzheimer’s Disease(AD)is a kind of senile neurodegenerative disease with hidden onset and related to cognitive ability.It develops from early mild cognitive impairment(EMCI)to late mild cognitive impairment(LMCI)and finally to AD.Therefore,the research of brain aging and senile neurological diseases is one of the hot topics in the cross fields of medicine,computer,biomedical engineering and so on.Brain is a nonlinear complex system,and complexity is a new index to explore brain.The researchers analyzed the brain complexity of AD patients in different modes and found that the brain complexity of AD patients was significantly lower than that of Normal Control(NC).However,the current research on brain aging mostly uses brain network technology and lacks the research on complexity,which leads to the lack of continuity research on complexity from aging to senile diseases.Permutation entropy(Per En)is a common complexity analysis method,which can describe the disorder and chaos degree of non-stationary signals.In this paper,Per En method was used to study the changes of brain complexity in patients with brain aging and AD from voxel and module perspectives.First,we use NKI-RS database to study the change curve of brain aging,then study the abnormal pattern of AD complexity,and make a comprehensive analysis with the existing literature on AD complexity.Finally,we use the complexity index to classify AD.The main research contents are as follows:(1)Regression and cluster analysis were used to study the changes of brain complexity during aging.In the process of aging,the aging time of each brain region and brain tissue is not completely consistent.This study used Per En as an index to explore the change pattern of brain complexity in the life cycle.Regression analysis showed that most brain regions showed an inverted U-shaped aging trend,with aging ages ranging from 25 to 51.Through cluster analysis of aging trend,four basic aging models were found.(2)Analysis of variance and gene information were used to study the changes of brain complexity in patients with AD.AD is a common senile disease.Researchers found that Apolipoprotein E(APOE)ε4 is the main genetic factor of AD.However,the relationship between APOE ε4 gene and brain complexity in AD patients is unknown.In this study,we combined Per En with gene to study the differences of complexity between NC,EMCI,LMCI and AD carriers and non carriers.The main differences were found in the superior frontal gyrus,middle occipital gyrus and parietal lobe.(3)The preferred reporting items for systematic reviews and meta analyses(PRISMA)was used to analyze the changes of brain complexity in different modes of AD.Although there are some researches on different modal complexity of AD,the research results are messy and lack of comprehensive review and analysis.This paper reviews the related research from 2000 to 2021.Objective to reveal the pattern of brain dysfunction in AD patients and explore whether complexity can be used as a biomarker of neurological impairment in AD.The results show that the EEG and brain magnetic signals of AD patients have been analyzed in previous studies,and the results of this study are basically consistent with those of previous studies: compared with NC,the signal complexity of AD patients is lower,and the differences are mainly distributed in the left parietal lobe,occipital lobe,right frontal lobe and temporal lobe,indicating that complexity may be used as a biomarker of neurological impairment in AD.(4)Combined with gene information and complexity index,the classification model of patients with Alzheimer’s disease based on Support Vector Machine(SVM)is realized.On the basis of the above research,using the complexity and gene information,training the classification model of AD based on SVM,the early intervention diagnosis of AD patients is explored.The classification results show that the accuracy of classification is significantly improved compared with the existing studies,which further shows that the complexity of comprehensive gene information may be used as a biomarker of neural function damage in AD.