Premium
IC‐P‐032: INFLUENCE OF NETWORK CONSTRUCTION METHODS ON PATH LENGTH VALUES IN ALZHEIMER'S DISEASE: A MULTI‐STUDY ANALYSIS OF MRI CONNECTIVITY STUDIES
Author(s) -
Dicks Ellen,
He Yong,
Beg Faisal,
Biessels Geert Jan,
Carmeli Cristian,
Chen Guangyu,
Çiftçi Koray,
Dai Zhengjia,
Van Dellen Edwin,
Dyrba Martin,
Friedman Eric J.,
Haan Willem,
Jalili Mahdi,
Jeong Jaeseung,
Jeong Yong,
Jiang Tianzi,
Knyazeva Maria G.,
Lee Dong Young,
Li Shi-Jiang,
Liu Yong,
McCarthy Paul,
Rodriguez Luis Peraza,
Phillips David,
Raamana Pradeep Reddy,
Reijmer Yael D.,
Seo Eun Hyun,
Sepulcre Jorge,
Soldan Anja,
Suckling John,
Taylor John-Paul,
Teng Gao-Jun,
Wink Alle Meije,
Barkhof Frederik,
Scheltens Philip,
Flier Wiesje M.,
Tijms Betty M.
Publication year - 2018
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2018.06.2096
Subject(s) - path length , functional connectivity , resting state fmri , path analysis (statistics) , path (computing) , psychology , pattern recognition (psychology) , computer science , audiology , medicine , artificial intelligence , neuroscience , machine learning , computer network , programming language
of DMN effective connectivity in AD, such as used models of hemodynamic effective connectivity along with t-tests to compare connections between groups. Dynamic causal modeling (DCM) was used to explore neuronal effective connectivity differences between healthy controls and a group of mild cognitive impairment (MCI) and AD patients. Methods:Resting state fMRI data were acquired on 3T GE 750 scanners at the University of Wisconsin-Madison and Medical College of Wisconsin using whole-brain simultaneous multi-slice imaging (8 bands, 72 slices, TR 802 ms, voxel size 2 mm isotropic). Participants included 11 controls (6 females, average age1⁄469.45 years) and 11 patients (5 females, average age1⁄469.82 years), 8 MCI and 3 AD, (Table 1) enrolled in the NIH-sponsored Alzheimer’s Disease Connectome Project. Following the Human Connectome Project minimal preprocessing pipeline , six motion parameters, motion parameters’ derivatives, and white matter and CSF signals were regressed out. DMN seed regions were extracted using coordinates (Table 2) in Van Dijk et al. (2010) to construct fully connected spectral DCMs. DCMs were the priors in a parametric empirical Bayes (PEB) model. Parameters in the second level Bayesian regression were group, sex, and mean centered age. Non-evident connections were removed via automated search. Results:The final PEB model included seven connections (Figure 1) which differed between patients and controls following thresholding of automated search results for parameters with a posterior probability of 95% of being non-zero. Connections for controls are depicted in Figure 2. Patients showed decreased self-inhibition, denoted by a positive value , in the posterior cingulate cortex (PCC), left and right hippocampal formation