Research themes
Contents
Functional architecture in health and disease
Our work on functional architecture primarily explores how cognitive function arises from the topographical organization of the cortex and on a secondary level, to what extent cortical organisation constrains functional processes (e.g. Pappas et al., 2020; Luppi et al 2022). The work encompasses the study of brain network behaviour evoked from tasks (e.g. Vatansever et al., 2016; Stamatakis et al., 2005) but also from resting state acquisitions (e.g. Pappas et al., 2019; Stamatakis et al., 2010).
Group-average matrices display the redundant (a) and synergistic (b) interactions between each pair of brain regions. Brain plots show the cortical distribution of the strongest redundant (blue) and synergistic (red) connections (thresholded to retain the top 5% of connections, for display purposes only). SOM, somatomotor network; VIS, visual network; SAL, salience/ventral attention network; DAN, dorsal attention network; LIM, limbic network; SUB, subcortical network. NeuroSynth term-based meta-analysis, relating the distribution of redundancy-to-synergy gradient across the brain to a cognitive gradient of cognitive domains, from lower-level sensorimotor processing to higher-level cognitive tasks.
The default mode network and cognition
The brain network that maintains a remarkable level of popularity is the default mode network (DMN). Cortical architecture studies place the default mode network (DMN) at the apex of several hierarchal cortical organisation schemes. Additionally, numerous reports describe DMN alterations not only in altered states of consciousness (Luppi et al., 2019; Stamatakis et al., 2010) but also in a broad spectrum of neuropsychiatric and neurodegenerative disorders such as depression (Pappas et al., 2020 -preprint), post-traumatic stress disorder, schizophrenia and Alzheimer’s disease. The midline brain structures that the DMN encompasses are known to respond in self-referential processing and although it was initially considered to be a task negative network (i.e. deactivates during tasks) studies from our group have shown DMN activity/connectivity during a wide range of tasks (Vatansever et al., 2018; 2017; 2015; Lu et al., 2021; 2022). An emerging theory is that activity in the DMN is neither uniquely intrinsic nor uniquely extrinsic and that it acts as (or is part of) a global neuronal workspace integrating information from a variety of sources to make sense of the world at present.
a Mean fractional occupancy (FO) values averaged across trials, grouped by the two categorical variables: conditions and states. For the box plots, the white dot/circle on each box indicates the mean, while the length of the box indicates quantiles with the middle bar showing the median. b Radar chart showing the correspondence between conditions and states. c Spectral information of the states in the alpha band. The asterisk on the heatmap indicates that the phase coherence (or power) in this state is significantly higher (yellow) or lower (blue) compared with the other states. The circular bundle plots below highlight the significant connectivities, with red and blue respectively signifying higher and lower significance for the connectivity.
Brain Dynamics and complexity
The human brain is the origin of a wide range of behaviours which are underpinned by a complex functional repertoire, developed upon a fixed structural anatomy (Pappas et al., 2020). The brain's functional repertoire varies over time, exhibiting dynamics that have recently been shown capable of differentiating awake from anaesthetised healthy volunteers and healthy volunteers from patients with disorders of consciousness (Coppola et al., 2022; Pappas et al., 2019). Furthermore, recent scientific theories of consciousness have emphasised the brain’s complexity as a crucial requirement for consciousness. Our research investigates the temporal and spatial aspects of the interaction between brain dynamics and complexity in supporting consciousness (Luppi et al., 2019). We employ entropy, as well as other measures of diversity or unpredictability of information content to quantify various aspects of brain function (Varley et al., 2020). We also use fractal properties (e.g. Luppi et al., 2021; Varley et al., 2020; Varley et al., 2020) to explore the insights that functional complexity can contribute to our understanding of both healthy and damaged brain function.
Correlation matrices between different complexity metrics for two different datasets. All entries along the diagonal have been removed. There are some typical patterns: the graph measures (LZ_Graph and Algebraic Connectivity) are both generally more highly correlated, as are LZC, SampEn and Hurst. The p-values ranged over many orders of magnitude from 10−2 to 10−20.
Neurobiology of consciousness
A significant challenge of contemporary neuroscience is to understand how the neurobiology of the human brain gives rise to conscious experience. One way to address this question is to identify changes in functional architecture that accompany changes in conscious state. We investigate consciousness alterations that occur through pharmacological interventions such as anaesthetic drugs (Varley et al., 2020a; Pappas et al., 2019; Stamatakis et al., 2010) or psychedelics (Luppi et al., 2021; Varley et al., 2020b) but also through trauma such as hypoxic-ischemic injuries or localised traumatic brain injuries (Varley et al., 2020c; Luppi et al., 2019). We have shown that, during propofol-induced anaesthesia, brain networks (particularly the DMN) exhibit altered connectivity as well as complexity and we have proposed these to be critical features of any candidate mechanistic description for loss of consciousness (Luppi et al., 2019; Stamatakis et al., 2010).
Brain maps of consciousness-related reductions in intrinsic connectivity contrast (ICC) and sample entropy and their overlap between and within two datasets of anaesthetised healthy volunteers and patients with disorders of consciousness (DOC).
Networks that demonstrate significantly reduced entropy with increasing sedation.
Neuromodulatory effects of brainstem nuclei (Ultra-High-Field MR Imaging)
Preclinical work has shown that projections from brainstem nuclei to the cortex are important for awareness and wakefulness. Despite this, no account exists of the functional modulatory relationships between these nuclei and cortical/subcortical networks, relevant for consciousness and associated disorders. We used 3T fMRI data from anaesthetised healthy volunteers and patients with disorders of consciousness (DoC) to characterize a brainstem functional connectivity signature of consciousness. This approach can potentially aid the identification of specific neurotransmitter systems as viable therapeutic targets for pharmacological interventions. Our findings so far support an important role for the dopaminergic system in consciousness (Spindler et al., 2001). To overcome limitations in existing 3T MRI imaging of the human brainstem, we are collecting high spatial/temporal resolution, ultra-high field 7T MRI structural/functional data. This work has a precision medicine frame of reference, targeting the creation of personalised neurotransmitter nuclei connectomes that will inform treatment options and facilitate the monitoring of treatment in a range of neurological and psychiatric conditions.
VTA connectivity (A) Healthy Controls (B) In a contrast of patients to these awake controls. (C) The posterior disconnection clusters from sedation and DoC datasets spatially overlapped. (D) Display of posterior region from B.
Neurobiology of language comprehension
Comprehending spoken language involves processing the sounds of words, their morphological structure and meaning and constructing higher level syntactic and semantic representations. These processes and representations are instantiated in an extensive, primarily left lateralised, frontal temporal and parietal set of brain areas connected by major white matter tracts such as the superior longitudinal and uncinate fasciculi. Coordinated activity between these regions may be critical in supporting comprehension of and responses to spoken stimuli, thus contributing to conscious awareness of speech. We use functional MRI to understand how attenuated consciousness impacts on coordination between brain areas and ultimately on language comprehension. Our cognitive neuroscience approach has implications for understanding language comprehension on a neural level as well as establishing the degree of preserved cognition at different computational levels (e.g. semantic processing, syntax) in patients with disorders of consciousness.
Changes in activity in the LIFG (top panel), LMTG (middle panel) and LITG (lower panel) in response to increasing sedation. * denotes significance at P < 0.05. Error bars show standard error. The relationship between changes in activity in the LIFG and plasma propofol concentrations is shown in the correlation plots to the right. Activity in the LIFG negatively correlates with plasma propofol concentration
The neural basis of cognitive, behavioural and emotional outcomes following traumatic brain injury
Many patients who survive traumatic brain injury suffer from neuropsychological sequelae such as memory loss, poor concentration and impulsivity. There is little direct data on the neurochemical systems that are involved in these cognitive deficits. We investigate the effects of cognitive enhancers such as methylphenidate on these compromised cognitive functions to understand the role of different neurotransmitter systems in cognitive deficits following TBI. Such pharmacological fMRI studies provide information on differential effects of these agents on activation patterns associated with improved cognitive performance, and provide a rational basis for selecting more targeted neurocognitive enhancers to treat neuropsychological sequelae following TBI (Dorer et al., 2018; Manktelow et al., 2017; Moreno-López et al., 2017). Along with the physical and cognitive sequelae, major depressive disorder is among the most frequent complications observed in patients following TBI. Patients experiencing depression after TBI have increased levels of anxiety, cognitive deficits and disability in comparison with those who do not develop depression. Our work in this area explores the relationship between severity of depressive symptoms and functional network architecture in patients with TBI (Pappas et al., 2020 -preprint; Moreno-López et al., 2016).
Within-group functional connectivity maps calculated from a seed in the subgenual cingulate cortex, found to be associated with depression scores in traumatic brain injury patients (TBI).
