Graph-theoretical measures provide translational markers of large-scale brain network disruption in human dementia patients and animal models of dementia

Stothart, George, Petkov, George, Kazanina, Nina, Goodfellow, Marc, Tait, Luke and Brown, Jon 2016. Graph-theoretical measures provide translational markers of large-scale brain network disruption in human dementia patients and animal models of dementia. International Journal of Psychophysiology 108 , p. 71. 10.1016/j.ijpsycho.2016.07.232

Full text not available from this repository.

Abstract

Introduction: Graph-theoretical measures of large-scale brain networks are sensitive to network disruption in a range of neurological disorders (Stam, 2014, Nat Revs Nscience). We have used graph theoretical methods to analyse EEG from human subjects with Alzheimer’s disease (AD, n = 21), mild cognitive impairment (MCI, n = 25) and healthy older adults (HOA, n = 26). We used the same methods to examine EEG from the CHMP2B transgenic mouse model of dementia in order to establish common translational markers of dementia across humans and animals. Methods: Resting-state, eyes-open EEG was recorded from medication-free human subjects and CHMP2B mice. 20 s artefact-corrected epochs were extracted and filtered in a range of frequency bands. Spectral analysis and functional connectivity matrices were created for each frequency band. These matrices were analysed using graph theoretical approaches to investigate the properties of the network. Results: Spectral power in the 2-5 Hz frequency band was significantly reduced in both the AD and MCI group when compared to HOA. It was also reduced in the CHMP2B mice compared to wild type. Network analyses revealed significant losses of heterogeneity and small-world topology in brain networks in both AD and MCI patients compared to HOA, and CHMP2B mice compared to wild type. A classifier derived from spectral and network measures was able to classify AD patients with 91% sensitivity, and successfully identified the MCI patients who later converted to AD. Conclusions: These data reveal that graph-theoretical measures of large-scale brain networks have translational value in bridging human and animal dementia research, and may have diagnostic value in predicting conversion from MCI to AD.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Cardiff University Brain Research Imaging Centre (CUBRIC) Psychology
Publisher:	Elsevier
ISSN:	0167-8760
Date of Acceptance:	October 2016
Last Modified:	15 Mar 2024 15:57
URI:	https://orca.cardiff.ac.uk/id/eprint/166679

Actions (repository staff only)

Edit Item