Format

Send to

Choose Destination
Comput Diffus MRI (2014). 2014;2014:55-64. Epub 2014 Dec 12.

Algebraic connectivity of brain networks shows patterns of segregation leading to reduced network robustness in Alzheimer's disease.

Author information

1
Imaging Genetics Center, Institute for Neuroimaging & Informatics, University of Southern California, Los Angeles, CA, USA madelaine.daianu@ini.usc.edu.
2
Imaging Genetics Center, Institute for Neuroimaging & Informatics, University of Southern California, Los Angeles, CA, USA neda.jahanshad@ini.usc.edu.
3
Imaging Genetics Center, Institute for Neuroimaging & Informatics, University of Southern California, Los Angeles, CA, USA talia.nir@ini.usc.edu.
4
Imaging Genetics Center, Institute for Neuroimaging & Informatics, University of Southern California, Los Angeles, CA, USA cassandra.leonardo@loni.usc.edu.
5
Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA jack.clifford@mayo.edu.
6
Department of Radiology, Medicine, and Psychiatry, University of California San Francisco, CA, USA michael.weiner@ucsf.edu.
7
Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA mBernstein@mayo.edu.
8
Imaging Genetics Center, Institute for Neuroimaging & Informatics, University of Southern California, Los Angeles, CA, USA pthomp@usc.edu.

Abstract

Measures of network topology and connectivity aid the understanding of network breakdown as the brain degenerates in Alzheimer's disease (AD). We analyzed 3-Tesla diffusion-weighted images from 202 patients scanned by the Alzheimer's Disease Neuroimaging Initiative - 50 healthy controls, 72 with early- and 38 with late-stage mild cognitive impairment (eMCI/lMCI) and 42 with AD. Using whole-brain tractography, we reconstructed structural connectivity networks representing connections between pairs of cortical regions. We examined, for the first time in this context, the network's Laplacian matrix and its Fiedler value, describing the network's algebraic connectivity, and the Fiedler vector, used to partition a graph. We assessed algebraic connectivity and four additional supporting metrics, revealing a decrease in network robustness and increasing disarray among nodes as dementia progressed. Network components became more disconnected and segregated, and their modularity increased. These measures are sensitive to diagnostic group differences, and may help understand the complex changes in AD.

KEYWORDS:

Alzheimer's disease; Fiedler value; algebraic connectivity; brain network; modularity

Supplemental Content

Full text links

Icon for PubMed Central
Loading ...
Support Center