Study Of Brain Morphology And Structural Networks For AD Patients Based On MRI

With the development of population aging, Alzheimer’s disease (AD)has become a common age-related neuro-degenerative diseases,seriously impacting on the patient’s daily life and inducing heavy socialload. Mild cognitive impairment (MCI), especially amnestic MCI (aMCI),has high risk to develop into AD. Early diagnosis, intervention and controlof aMCI may reduce the shifting rate from aMCI to AD, which hasimportant clinicial and social significance.This study investigated the pathological alterations and mechanismsof AD and aMCI based on their magnetic resonance imaging (MRI) data.In particular, this study examined the brain tissue atrophy under thepathological state by voxel-based morphological analysis, and investigatedthe alterations of structural brain networks of AD and aMCI patients basedon their diffusion tensor imaging (DTI) from brain network topology andpair-wise white matter connectivity between brain regions. The correlationbetween the brain network/connectivity parameters and cognitive functionslevel was further studied, in order to find network features related to ADdisease and thus provide a potential biomarker for early clinical diagnosisand prediction of the progress of AD and aMCI patients.We found that many brain regions, such as the hippocampus, frontallobe, temporal lobe and insula showed significant atrophy in AD and aMCIpatients, especially in AD patients. The characteristic path length of ADpatents’ brain structural networks significantly increased and the networkstrength significantly decreased. There were significant group effects in thenodal characteristics of several brain regions, such as the thalamus,entorhinal cortex, heschl gyrus, precuneus, superior parietal gyrus, insula and hippocampus. The hub regions of the three groups were in highconsistency, though with significantly different nodal betweenness.Meanwhile, there were less hub regions in the frontal lobe in AD groupthan aMCI and normal control (NC) groups. From the perspective ofconnectivity, the FA weights of many connectivity, especially theconnectivity of long fiber length, were significantly changed in ADpatients. In addition, correlation analysis showed that some parameters ofbrain volumes and network had significant correlation to cognitivefunction level in the AD, aMCI and NC groups.The results of brain morphology and structural network shows thatAD is a disease related to neuron loss and white matter lesions, and thestructure changes in brain volumes and white matter architecture of aMCIpatients are in a mid-stage between AD patients and normal aging. Thedecreased FA values of connectivity, especially of connectivity with longfiber length, implies that AD is related to the reduce of connectivitybetween brain regions. Some parameters, such as brain volumes andnetwork characteristics are significantly correlated with cognitive functionlevel, implying that the study of morphology and structural brain networkscan reveal the extent of the damage in patients. Thus, these brainparameters may be potential biomarkers to provide early diagnosis andprediction of AD and aMCI patients.