Altered default mode network connectivity in alzheimer's disease-A resting functional MRI and bayesian network study

• Published in: Human brain mapping Vol. 32; no. 11; pp. 1868 - 1881
• Main Authors: Wu, Xia, Li, Rui, Fleisher, Adam S., Reiman, Eric M., Guan, Xiaoting, Zhang, Yumei, Chen, Kewei, Yao, Li
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc., A Wiley Company 01.11.2011; Wiley-Liss; John Wiley & Sons, Inc
• Subjects: Aged; Algorithms; Alzheimer Disease - physiopathology; Alzheimer's disease; Bayes Theorem; Bayesian analysis; Biological and medical sciences; biomarker; biomarkers; Brain Mapping; Cerebral Cortex - physiopathology; Cortex (parietal); Cortex (prefrontal); Cortex (temporal); Data Interpretation, Statistical; Degenerative and inherited degenerative diseases of the nervous system. Leukodystrophies. Prion diseases; effective connectivity; Female; fMRI; functional connectivity; Functional magnetic resonance imaging; Hippocampus; Hippocampus - physiopathology; Humans; Image Processing, Computer-Assisted; Investigative techniques, diagnostic techniques (general aspects); Magnetic Resonance Imaging; Male; Medical sciences; Middle Aged; Nerve Net - physiopathology; Nervous system; Neural networks; Neurodegenerative diseases; Neurology; Principal Component Analysis; Radiodiagnosis. Nmr imagery. Nmr spectrometry; resting state; fMRI; Nervous system diseases; resting state; functional connectivity; Radiodiagnosis; Alzheimer disease; Central nervous system disease; effective connectivity; Degenerative disease; biomarker; Nuclear magnetic resonance imaging; Cerebral disorder
• ISSN: 1065-9471; 1097-0193; 1097-0193
• DOI: 10.1002/hbm.21153

A number of functional magnetic resonance imaging (fMRI) studies reported the existence of default mode network (DMN) and its disruption due to the presence of a disease such as Alzheimer's disease (AD). In this investigation, first, we used the independent component analysis (ICA) technique to confirm the DMN difference between patients with AD and normal control (NC) reported in previous studies. Consistent with the previous studies, the decreased resting‐state functional connectivity of DMN in AD was identified in posterior cingulated cortex (PCC), medial prefrontal cortex (MPFC), inferior parietal cortex (IPC), inferior temporal cortex (ITC), and hippocampus (HC). Moreover, we introduced Bayesian network (BN) to study the effective connectivity of DMN and the difference between AD and NC. When compared the DMN effective connectivity in AD with the one in NC using a nonparametric random permutation test, we found that connections from left HC to left IPC, left ITC to right HC, right HC to left IPC, to MPFC and to PCC were all lost. In addition, in AD group, the connection directions between right HC and left HC, between left HC and left ITC, and between right IPC and right ITC were opposite to those in NC group. The connections of right HC to other regions, except left HC, within the BN were all statistically in‐distinguishable from 0, suggesting an increased right hippocampal pathological and functional burden in AD. The altered effective connectivity in patients with AD may reveal more characteristics of the disease and may serve as a potential biomarker. Hum Brain Mapp, 2011. © 2011 Wiley‐Liss, Inc.