A5090246172- Functional Brain Connectivity Studies
- Neural dynamics and brain function
- Neuroscience and Music Perception
- Neural and Behavioral Psychology Studies
- Epilepsy research and treatment
- Neuroscience and Neuropharmacology Research
- EEG and Brain-Computer Interfaces
- Child and Animal Learning Development
- Advanced MRI Techniques and Applications
University of Cambridge
2015-2022
Medical Research Council
2015-2022
MRC Cognition and Brain Sciences Unit
2015-2022
Cambridge University Hospitals NHS Foundation Trust
2020
Imperial College London
2011
Brain function can be conceived as a hierarchy of generative models that optimizes predictions sensory inputs and minimizes "surprise." Each level the makes neural events at lower in hierarchy, which returns prediction error when these expectations are violated. We tested generalization this hypothesis to multiple sequential deviations, we identified most likely organization network accommodates deviations temporal structure stimuli. Magnetoencephalography healthy human participants during...
We propose that sensory inputs are processed in terms of optimised predictions and prediction error signals within hierarchical neurocognitive models. The combination non-invasive brain imaging generative network models has provided support for frontotemporal interactions oddball tasks, including recent identification a temporal expectancy signal acting on prefrontal cortex. However, these studies limited by the need to invert magnetoencephalographic or electroencephalographic sensor...
The localization of intracranial electrodes is a fundamental step in the analysis invasive electroencephalography (EEG) recordings research and clinical practice. conclusions reached from these rely on accuracy electrode relationship to brain anatomy. However, currently available techniques for localizing magnetic resonance (MR) and/or computerized tomography (CT) images are time consuming limited particular types or shapes. Here we present iElectrodes, an open-source toolbox that provides...
The clinical syndromes caused by frontotemporal lobar degeneration are heterogeneous, including the behavioural variant dementia (bvFTD) and progressive supranuclear palsy. Although pathologically distinct, they share many behavioural, cognitive physiological features, which may in part arise from common deficits of major neurotransmitters such as γ-aminobutyric acid (GABA). Here, we quantify GABAergic impairment its restoration with dynamic causal modelling a double-blind placebo-controlled...
To bridge the gap between preclinical cellular models of disease and in vivo imaging human cognitive network dynamics, there is a pressing need for informative biophysical models. Here we assess dynamic causal (DCM) cortical responses, as generative magnetoencephalographic observations during an auditory oddball roving paradigm healthy adults. This induces robust perturbations that permeate frontotemporal networks, including evoked 'mismatch negativity' response transiently induced...
The multiple demand (MD) system is a network of fronto-parietal brain regions active during the organization and control diverse cognitive operations. It has been argued that this activation may be nonspecific signal task difficulty. However, here we provide convergent evidence for causal role MD in “simple task” automatic auditory change detection, through impairment top-down mechanisms. We employ independent structure-function mapping, dynamic modeling (DCM), frequency-resolved functional...
Choosing between equivalent response options requires the resolution of ambiguity. One could facilitate such decisions by monitoring previous actions and implementing transient or arbitrary rules to differentiate options. This would reduce entropy chosen actions. We examined voluntary action during magnetoencephalography, identifying spatiotemporal correlates stimulus- choice-entropy. Negative correlations frontotemporal activity past trials were observed after participants' responses,...
The exploration-exploitation trade-off is at the heart ofany sequential decision-making process in uncertainenvironments, such as foraging for food [1] or learningto cycle. This characterises balancebetween reaping benefits of a known solution(which may not be optimal) and continuing tosearch hope better solutions. It difficult todirectly infer measure this from behaviouraldata humans (or animals) on trial-by-trial basis.Here, we use set reinforcement learning (RL) algo-rithms ideal actor...
Abstract To bridge the gap between preclinical cellular models of disease and in vivo imaging human cognitive network dynamics, there is a pressing need for informative biophysical models. Here we assess dynamic causal (DCM) cortical responses, inverted to magnetoencephalographic observations during an auditory oddball roving paradigm healthy adults. This induces robust perturbations that permeate frontotemporal networks, including evoked ‘mismatch negativity’ response transiently induced...
Abstract Background Magnetoencephalography (MEG) is a promising tool for experimental medicine in dementia, and core technology the Dementias Platform UK. It quantifies vivo human network synaptic physiology, with high‐dimensional data at millisecond time‐scale. Its proven sensitivity to neurodegenerative disease has applications diagnostics early detection, mechanistic studies of pathogenesis, biomarkers accelerate clinical trials evidence target engagement. Here, we build on recent work...