Neural Network Study Of Alzheimer Disease Based On Diffusion Tensor Imaging

As the growing of elder population ratio in China,Alzheimer Disease(AD)has become the largest disease that brings burden to our society.More and more economic issue as well as social issue are rising because of AD.Research on AD are important and urgent because its high morbidity,fatality rate and low curable possibility.In the past,diagnose of AD were mainly depend on looking for nerve tangles and senile plaques in dead people's brain and there was no way of AD definite diagnosis for the living,which made early diagnosis and early intervention are important.Course of AD can be divided into pre-clinical AD,Mild Cognitive Impairment(MCI)and phase of dementia.But,the intervening mode and big expense makes current examination not feasible for large-scale screening.Appearing of Magnetic Resonance Imaging(MRI)technology provides a non-invasive method for the detection of neurological disease.It can reflect the structure,function,neural junction,blood oxygen consumption level response to different imaging sequence.This study is based on MRI and Diffusion Tensor Imaging(DTI),along with the research orientation of brain network program.With the help of probabilistic nerve fiber tracking algorithm and multigraph segmentation technique,this study built the nerve connected network of whole brain,came up with a new theory of Cognitive Control Related Network(CCRN)based on hub and spoke model.Making quantitative analysis of CCRN and whole brain network with graph theory attributes like local nodal degree,local nodal efficiency and global small world property.Statistical tests are also employed to figure out the difference between groups.The relationship between various of graph theory attributes and Mini-Mental State Examination(MMSE)score were analyzed in order to extract the graph features that can describe the course of AD.Physiology meaning of these graph features are also talked.Finally,a systematized hierarchy AD early warning method was built base on these graph features and machine learning method.The result of this study showed the huge alteration of connectivity in whole brain network and CCRN responding to the progress of AD.With the help of graph theory analysis and machine learning method,description of AD can be particular and explicable.This study provided a new view and broader technological means for AD research as well as other neuron degenerative diseases.