Connectome-Based Modelling Reveals Ketamine's Modulatory Effects on Thalamocortical Connectivity
Poster No:
1672
Submission Type:
Abstract Submission
Authors:
Andreea Diaconescu1, Zheng Wang2, Milad Soltanzadeh1, Davide Momi3, Andrew Clappison4, André Schmidt5, Franz Vollenweider6, John Griffiths1
Institutions:
1University of Toronto, Toronto, Ontario, 2Centre for Addiction and Mental Health, Toronto, Ontario, 3CAMH, Toronto, Ontario, 4University of Ottawa, Ottawa, Ontario, 5University of Basel, Department of Clinical Research (DKF), University Psychiatric Clinics (UPK), Tr, Basel, Basel, 6University Hospital of Psychiatry, Zurich, Ontario
First Author:
Co-Author(s):
André Schmidt
University of Basel, Department of Clinical Research (DKF), University Psychiatric Clinics (UPK), Tr
Basel, Basel
University of Basel, Department of Clinical Research (DKF), University Psychiatric Clinics (UPK), Tr
Basel, Basel
Introduction:
Ketamine, an NMDA receptor antagonist, is a pharmacological model for schizophrenia as it induces schizophrenia-like cognitive impairments at subanesthetic doses in healthy individuals (1). Its antagonism of NMDA receptors, especially in the prefrontal cortex, indirectly boosts dopamine release (2). The increase in dopamine levels following ketamine administration can enhance the activity of D1 receptors, which are predominantly excitatory. This can affect various cognitive processes and is particularly relevant in the context of schizophrenia spectrum disorders (SSD), where D1 receptor dysregulation is implicated (3). To understand ketamine's effect on D1/D2 receptors and thalamocortical connectivity, we modelled the effects of ketamine during sensory learning using the auditory mismatch negativity (MMN) paradigm.
Methods:
To assess ketamine's effect on brain dynamics in an auditory MMN, we re-analyzed a previously published EEG dataset using a placebo-controlled, crossover design (4). We employed a neural mass model (5) in a cortico-striato-thalamo-cortical (CSTC) circuit framework. This incorporated the 7 Yeo networks with 200 subdivisions (6) as a structural prior. The CSTC circuit (Fig. 1) included subcortical and cortical-subcortical connections. Each node comprised three neuronal populations: pyramidal, excitatory, and inhibitory, modeled using the Jansen-Rit (JR) approach.
We simulated auditory responses by introducing a step stimulus into auditory-related regions and running the JR neural mass model with the CSTC circuit to generate EEG signals using a leadfield matrix. The model was implemented in the WhoBPyT Python library, allowing optimization to empirical EEG data using mean-square error as a cost function and PyTorch for automatic gradient calculation, updating node and population connections, and the leadfield matrix until parameter convergence (7).
The optimized parameters, particularly connection gains within nodes and across the CSTC circuit, were evaluated to compare ketamine's effects against placebo during the MMN paradigm.
We simulated auditory responses by introducing a step stimulus into auditory-related regions and running the JR neural mass model with the CSTC circuit to generate EEG signals using a leadfield matrix. The model was implemented in the WhoBPyT Python library, allowing optimization to empirical EEG data using mean-square error as a cost function and PyTorch for automatic gradient calculation, updating node and population connections, and the leadfield matrix until parameter convergence (7).
The optimized parameters, particularly connection gains within nodes and across the CSTC circuit, were evaluated to compare ketamine's effects against placebo during the MMN paradigm.
Results:
Partial least squares (PLS) analysis was used to assess condition-by-drug interactions based on these model-derived parameters. Firstly, we found significant group differences (p=0.03) between the ketamine and placebo conditions in the "standard tone" response. Specifically, the connection gain from the pyramidal (P) to inhibitory (I) populations (C3) in the JR model and the thalamus to D1 connection gain were notably higher in the placebo condition. This implies more pronounced inhibition within nodes and stronger thalamic inhibition across nodes in the placebo compared to the ketamine condition. Secondly, PLS analysis showed a significant difference (p=0.04) in "deviant tone" responses between the ketamine and placebo conditions, with the ketamine condition exhibiting increased D2 connection gains in temporal, cingulate, and medial prefrontal regions, but decreased gains in parietal and dorsal middle cingulate areas (Fig 2A). Additionally, a substantial difference (p=0.0002) was found in thalamocortical connectivity during "standard tone" responses, with the posterior cuneus and anterior cingulate showing higher connectivity following ketamine administration, while the temporal, parietal, and medial prefrontal regions had increased connectivity under placebo (Fig. 2B).
Conclusions:
These findings underscore significant changes in brain connectivity patterns, specifically in thalamus to D1 and cortex to D2 connections, under the influence of ketamine compared to placebo. This indicates that ketamine markedly alters neural pathways, especially influencing cortical intra-connections and those between the cortex and thalamus, providing understanding into the neural underpinnings of schizophrenia spectrum disorders.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia)
Modeling and Analysis Methods:
EEG/MEG Modeling and Analysis 1
Novel Imaging Acquisition Methods:
EEG
Perception, Attention and Motor Behavior:
Perception: Auditory/ Vestibular
Physiology, Metabolism and Neurotransmission :
Pharmacology and Neurotransmission 2
Keywords:
Computational Neuroscience
Electroencephaolography (EEG)
Glutamate
Learning
Perception
Schizophrenia
Thalamus
1|2Indicates the priority used for review
Provide references using author date format
2. Krystal, J. H. et al. (1994) 'Subanesthetic Effects of the Noncompetitive NMDA Antagonist, Ketamine, in Humans: Psychotomimetic, Perceptual, Cognitive, and Neuroendocrine Responses'. Arch. Gen. Psychiatry 51, 199–214.
3. Stone, J. M., Morrison, P. D. & Pilowsky, L. S. (2007) 'Glutamate and dopamine dysregulation in schizophrenia–a synthesis and selective review'. J Psychopharmacol 21, 440–52 .
4. Schmidt, A. et al. (2012) 'Mismatch Negativity Encoding of Prediction Errors Predicts S-ketamine-Induced Cognitive Impairments'. Neuropsychopharmacology 37, 865–875.
5. Jansen, B. H. & Rit, V. G. (1995) 'Electroencephalogram and visual evoked potential generation in a mathematical model of coupled cortical columns'. Biol. Cybern. 73, 357–66.
6. Yeo, B. T. et al. (2011) 'The organization of the human cerebral cortex estimated by intrinsic functional connectivity'. J. Neurophysiol. 106, 1125–1165.
7. Griffiths, J. D. et al. (2022) 'Deep Learning-Based Parameter Estimation for Neurophysiological Models of Neuroimaging Data'. bioRxiv. Preprint at https://doi.org/10.1101/2022.05.19.492664.