Dynamical complexity of large-scale neurocognitive networks in healthy and pathological brain states

  • Thomas Alderson

Student thesis: Doctoral Thesis


Contemporary theories suggest that the brain operates in a metastable
regime of dynamics in which the tendencies for local areas to integrate
and segregate is simultaneously realised. Current theoretical and empirical
observations suggest that this behaviour occurs spontaneously through the
interaction of local dynamics with underling anatomical connectivity. The
metastable regime likely confers important behavioural qualities through
the flexible coupling and uncoupling of distributed cortical regions into
context-dependent neurocognitive networks. Thus, one of the principle
goals of neuroscience, to understand how structure and dynamics interact
to generate cognition, may be realised by leveraging the metastable
regime of dynamics to link across the interrelated domains of structure,
function, and cognition. Accordingly, the proposed approach, grounded in
dynamical systems theory, neuroimaging, and theoretical computer modelling,
aims to explore how: (1) complex metastable neural dynamics are
modulated by cognitive state; (2) structural connectivity confers cognitive
flexibility on a fixed network topology through metastable neural dynamics;
(3) structural disconnection impacts metastable neural dynamics and how
this relates to cognitive performance.
The thesis presents findings from three studies. The first uses the theoretical
framework of metastable coordination dynamics to explore how cognition
arises from the dynamic assembly of local areas into neurocognitive
networks. Previous work has suggested that the probability of transitioning
between network states is maximised when subjects are not explicitly engaged
in a task. Contrary to expectations, metastability between networks
was higher during task engagement than during periods of ‘cognitive rest’.
Task-based reasoning was characterised by dynamic stability in sensory regions
and dynamic flexibility in regions devoted to cognitive control. Critically,
this dynamic flexibility appeared to confer superior problem solving
ability in tests of fluid intelligence.
The second study leverages an example of incipient neurodegeneration,
mild cognitive impairment (MCI), to test the essential proposition that
cognitive deficits are linked to structural disconnection in the brain’s largescale
network architecture. Accordingly, this study examines the structural
connectivity between thalamus and key regions of the cortex implicated
in ‘cognitive rest’: the default mode network (DMN). Abnormal structural
connectivity and altered patterns of causation were identified in this
‘thalamo-DMN’ loop and, crucially, these deficits were linked to memory
recall. Taken together, these findings provide new insight into the causal
pathways underlying DMN dysfunction in MCI and Alzheimer’s disease
(AD) and provides preliminary evidence that AD represents a failure of
circulating information consistent with its status as a ‘disconnection syndrome’.
The third and final study uses a joint theoretical and empirical approach
to examine how dynamics and cognitive ability are shaped by the macroscopic
connectivity of the brain. Accordingly, this study investigates the
asymptotic decline of neural metastability in an example of structural disconnection, AD, and its prodrome, MCI. Whole-brain computer modelling
mechanistically linked reduced metastability to anatomical disconnection.
Moreover, metastability was linked to features of the brain’s structural
topology. Crucially, empirical estimates of metastability were linked to
global cognitive performance. Taken together, these findings suggest a critical
linkage between metastability, cognition, and network topology in the
damaged or diseased brain.
Overall, these three studies provide insight into the dynamic principles
by which cognitive architecture is organised and suggest that the metastable
regime of dynamics offers considerable potential as a theoretical and conceptual
framework for linking structure, function, and cognition in the human
Date of AwardAug 2019
Original languageEnglish
SupervisorLiam Maguire (Supervisor) & Damien Coyle (Supervisor)


  • Metastability, Resting state, fMRI, DTI, Neural dynamics, Whole-brain computer modelling
  • Resting state
  • fMRI
  • DTI
  • Neural dynamics
  • Whole-brain computer modelling

Cite this