-
Tytuł:
-
Detecting abnormal connectivity in schizophrenia via a joint directed acyclic graph estimation model.
-
Autorzy:
-
Zhang G; Department of Biomedical Engineering, Tulane University, New Orleans, LA 70118, USA.
Cai B; Department of Biomedical Engineering, Tulane University, New Orleans, LA 70118, USA.
Zhang A; New York State Psychiatry Institute and Department of Psychiatry, Columbia University, New York, NY 10032, USA.
Tu Z; UBTECH Sydney Artificial Intelligence Centre, The University of Sydney, NSW 2006, Australia.
Xiao L; School of Information Science and Technology, University of Science and Technology of China, Hefei, Anhui 230052, China.
Stephen JM; Mind Research Network a division of Lovelace Biomedical Research Institute, Albuquerque, NM, USA.
Wilson TW; Institute for Human Neuroscience, Boys Town National Research Hospital, 14090 Mother Teresa Ln, Boys Town, NE 68010, USA.
Calhoun VD; Tri-institutional Center for Translational Research in Neuroimaging and Data Science (TReNDS) Georgia State University, Georgia Institute of Technology, Emory University, Atlanta, GA 30030 USA.
Wang YP; Department of Biomedical Engineering, Tulane University, New Orleans, LA 70118, USA. Electronic address: .
-
Źródło:
-
NeuroImage [Neuroimage] 2022 Oct 15; Vol. 260, pp. 119451. Date of Electronic Publication: 2022 Jul 14.
-
Typ publikacji:
-
Journal Article; Research Support, N.I.H., Extramural; Research Support, U.S. Gov't, Non-P.H.S.
-
Język:
-
English
-
Imprint Name(s):
-
Original Publication: Orlando, FL : Academic Press, c1992-
-
MeSH Terms:
-
Schizophrenia*/diagnostic imaging
Brain/diagnostic imaging ; Brain Mapping/methods ; Case-Control Studies ; Humans ; Magnetic Resonance Imaging/methods
-
Grant Information:
-
R56 MH124925 United States MH NIMH NIH HHS; R01 GM109068 United States GM NIGMS NIH HHS; R01 EB020407 United States EB NIBIB NIH HHS; R01 MH103220 United States MH NIMH NIH HHS; R01 MH104680 United States MH NIMH NIH HHS; R01 MH107354 United States MH NIMH NIH HHS
-
Contributed Indexing:
-
Keywords: Brain; Connectivity analysis; Functional imaging; Probabilistic and statistical methods; fMRI analysis
-
Entry Date(s):
-
Date Created: 20220716 Date Completed: 20220816 Latest Revision: 20230417
-
Update Code:
-
20240104
-
DOI:
-
10.1016/j.neuroimage.2022.119451
-
PMID:
-
35842099
-
Functional connectivity (FC) between brain region has been widely studied and linked with cognition and behavior of an individual. FC is usually defined as the correlation or partial correlation of fMRI blood oxygen level-dependent (BOLD) signals between two brain regions. Although FC has been effective to understand brain organization, it cannot reveal the direction of interactions. Many directed acyclic graph (DAG) based methods have been applied to study the directed interactions but their performance was limited by the small sample size while high dimensionality of the available data. By enforcing group regularization and utilizing samples from both case and control groups, we propose a joint DAG model to estimate the directed FC. We first demonstrate that the proposed model is efficient and accurate through a series of simulation studies. We then apply it to the case-control study of schizophrenia (SZ) with data collected from the MIND Clinical Imaging Consortium (MCIC). We have successfully identified decreased functional integration, disrupted hub structures and characteristic edges (CtEs) in SZ patients. Those findings have been confirmed by previous studies with some identified to be potential markers for SZ patients. A comparison of the results between the directed FC and undirected FC showed substantial differences in the selected features. In addition, we used the identified features based on directed FC for the classification of SZ patients and achieved better accuracy than using undirected FC or raw features, demonstrating the advantage of using directed FC for brain network analysis.
(Copyright © 2022. Published by Elsevier Inc.)
