Research On Predicting The Course Of Alzheimer’s Disease Based On The Relationship Between Brain Structure And Function

In the study of Alzheimer’s disease(AD),researchers have found characteristic changes in the topology of functional brain networks in AD patients.Additionally,researchers have found a close relationship between structural brain networks and functional brain networks.By studying and modeling the structural brain network,we can predict the characteristics and performance of the functional brain network.However,functional brain imaging techniques have limitations and drawbacks,such as noise,body movements,respiration,and heart rate,which can interfere with imaging results and lead to large errors in AD progression prediction.In contrast,brain structural imaging does not have these issues.Therefore,this paper aims to use the relationship between brain structure and function to predict AD progression,leveraging the advantages of brain structural imaging to reduce errors caused by external factors.The main research contents of this paper are as follows:(1)A novel graph deep learning model is proposed for the study of the brain’s structuralfunctional relationship.The model first calculates the shortest path between each node in the structural brain network as the baseline for predicting the strength of connections between nodes in the functional brain network.Then,the final prediction result is obtained by adjusting this baseline using a graph auto-encoder.The model combines network communication features and deep learning abilities,and extends the link prediction method to weighted graphs.Experimental results show that the proposed model performs significantly better than traditional methods based on structural similarity in predicting brain structural connections.(2)The use of brain structural connections is implemented for AD progression prediction.First,the functional brain network is analyzed to obtain highly correlated brain regions and topological features with AD.Then,the AD progression prediction is completed by combining the predicted functional brain network,brain function topology features,and classification model.This paper successfully constructs a framework for AD progression prediction based on the relationship between brain structure and function,and preliminary experimental results are promising.(3)Designed and implemented a AD progression prediction system based on structured brain networks.