Research And Application For The Memory Imitation Of The Brain

Cognitive and memory disorders bring great economic burden and time cost to many families,and at the same time increase the risk of other diseases.Therefore,this thesis needs to study the brain of people with cognitive and memory dysfunction and conduct simulation experiments on their memory function.This thesis is based on functional and structural imaging data from alzheimer's disease subjects and healthy subjects.After preprocessing,the corresponding brain structure network and brain function network are constructed,and the two networks are analyzed by using graph theory.Then from two experimental group brain activity can be derived from the signal(ALFF signal,f ALFF signal,Reho signal)analysis of the abnormal brain areas,will be abnormal brain regions associated with attribute topology,using ebbinghaus forgetting curve and network efficiency,put forward suitable for cognitive dysfunction in patients with memory curve,then the simulation experiments by using attribute topology.Finally,according to the machine learning algorithm,functional network matrix,functional network topology attributes,brain activity signals were used as feature inputs to train and generate the auxiliary diagnosis model of cognitive and memory dysfunction.The main achievements of this thesis are as follows:(1)In this thesis,the corresponding network was constructed through the structural and functional image data of the experimental subjects,and it was found that the people with cognitive and memory dysfunction had different degrees of abnormalities in the global network efficiency,local network efficiency,small-world attribute,compatibility,hierarchy and modularity of the network attributes of the imitation brain network.(2)This thesis found that keep the mental labor is beneficial to maintain the brain structure,reduce the disease of brain structure damage and slow cognitive memory dysfunction in patients with brain network variation,also found that chronic insomnia can destroy the brain structure,resulting in the network topological properties on degradation and some early symptoms of cognitive memory dysfunction.(3)In this thesis,brain activity signals(ALFF signal,f ALFF signal and Reho signal)were extracted from f MRI data,and significant differences were found in patients' medial temporal lobe memory system,prefrontal lobe,diencephalon(thalamus),amygdala,cerebellum,Meynert basal ganglia and limbic striatum compared with healthy people.(4)In this thesis,combine the topology attributes of brain regions and network involved in cognitive memory function with the theory of attribute topology,and proposed the forgetting curve suitable for healthy people and people with cognitive memory impairment.In addition,combine the memory model based on attribute topology and memory brain region,and conducted a simulation comparison on the memory forgetting process of healthy people and people with cognitive memory dysfunction.At the same time visual memory forgetting process.(5)In this thesis,ALFF signal,f ALFF signal,Reho signal,brain network topology attributes and brain function network matrix were used as input to train the auxiliary diagnosis model of cognitive and memory dysfunction.After continuous iterative optimization,the final accuracy of the model was up to 99.4%.