Network control energy within the salience network underlies multi-task switching
Poster No:
474
Submission Type:
Abstract Submission
Authors:
Johannes Wiesner1, Anastasia Benedyk1, Jamila Andoh1, Maximilian Lueckel2, Anais Buhl1, Linden Parkes3, Lorenzo Caciagli4, Xiaosong He5, Dani Bassett6, Emanuel Schwarz1, Andreas Meyer-Lindenberg1, Heike Tost1, Urs Braun1
Institutions:
1Central Institute of Mental Health, Mannheim, Baden-Württemberg, 2Neuroimaging Center (NIC), Focus Program Translational Neuroscience, Johannes Gutenberg University, Mainz, Germany, 3Rutgers University, Philadelphia, PA, 4University of Bern, Bern, Bern, 5University of Science and Technology of China, Hefei, Anhui, 6UPenn, Philadelphia, PA
First Author:
Co-Author(s):
Maximilian Lueckel
Neuroimaging Center (NIC), Focus Program Translational Neuroscience, Johannes Gutenberg University
Mainz, Germany
Neuroimaging Center (NIC), Focus Program Translational Neuroscience, Johannes Gutenberg University
Mainz, Germany
Introduction:
Network Control Theory (NCT) is able to model the brain as a dynamical system that transitions through various discrete states. While previous research has focused on the identification of brain networks that are tightly coupled to certain behavioral domains, less research has been done on addressing the question which networks facilitate switching between these bio-behavioral domains (Insel et al., 2010). Here, we used a Multi-Task control theory framework in order to identify a set of regions of interests (ROIs) that is relatively more involved in steering the brain into any specific target state than the rest of the brain. We then investigated behavioral, neuromodulatory, topological and behavioral correlates of these ROIs leading to possible implications for patients in the mood-psychosis spectrum.
Methods:
We used task fMRI and DWI data from the HCP dataset from 763 subjects that had completed all fMRI tasks and all tasks of the NIH toolbox (Glasser et al., 2013). SPM12 was used to compute T-maps for all conditions of interest ("Working Memory", "Emotional Regulation", "Reward Processing", "Social Processing"). DWI data was reconstructed in DSI-Studio using q-space diffeomorphic reconstruction (Yeh & Tseng, 2011). A Glasser Atlas with additional subcortical regions from the Tian Atlas (Glasser et al., 2016; Tian et al., 2020) was used to both parcellate the T-maps and to obtain structural connectivity matrices using deterministic fiber-tracking. For each subject, optimal control energy was computed for all possible state-to-state-transitions using previously described methods (Braun et al., 2021). Random intercept models were set up for each brain region to obtain both estimates for the average energy input across the entire sample and for each individual subject. Otsu's method was used to obtain ROIs with particularly high energy inputs (Otsu, 1979). We correlated our statistical image with the salience network map derived from neurosynth. To investigate possible links to cognitive functioning, subjects' control energy inputs of our identified network were correlated with relevant measures of fluid intelligence using univariate and multivariate methods. Finally, we tested for differences in receptor densities for each of the major ascending neuromodulatory systems as well as for modal and average controllability differences between our ROIs and the rest of the brain.
Results:
We identified an extended network of regions that showed a positive correlation with the salience network map from neurosynth (r = 0.16, p < 0.01). ROIs were also significantly correlated with higher D2 receptor density (U = 20507, p < 0.01), higher modal (W = 0, p < 0.01), and lower average controllability values (W = 2692, p < .01). Finally, higher energy in the ROIs was significantly correlated with worse performance in the Card Sorting test (ρ = -.07, p < .05) and the Flanker Test (ρ = -.08, p < .05).
·Statistical image with T-values derived from mixed model analyses (ROIs are outlined with black contours)
Conclusions:
Our results support the hypothesis that the salience network plays a crucial role in the control of brain state transitions. The finding that our ROIs are associated with higher D2 receptor expression density supports previous works stating that these receptors are important for facilitating segregation in brain-state dynamics (Shine et al., 2019). This is further supported by the finding of higher modal controllability in the identified network. These findings, and the identified link between control energy and cognitive functioning, highlights the salience network as a promising target for further investigation in a larger clinical sample with a particular focus on patients in the mood-psychosis spectrum.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Anatomy and Functional Systems
Transmitter Systems
Novel Imaging Acquisition Methods:
Multi-Modal Imaging 2
Keywords:
Cognition
Dopamine
FUNCTIONAL MRI
Limbic Systems
Neurotransmitter
Psychiatric Disorders
RECEPTORS
Schizophrenia
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
Other - Brain Dynamics
1|2Indicates the priority used for review
Provide references using author date format
Glasser, M. F. (2016), ‘A multi-modal parcellation of human cerebral cortex’, Nature, 536(7615), 171–178
Glasser, M. F. (2013), ‘The minimal preprocessing pipelines for the Human Connectome Project’, NeuroImage, 80, 105–124
Otsu, N. (1979), ‘A Threshold Selection Method from Gray-Level Histograms’, IEEE Transactions on Systems, Man, and Cybernetics, 9(1), 62–66
Tian, Y. (2020), ‘Topographic organization of the human subcortex unveiled with functional connectivity gradients’, Nature Neuroscience, 23(11), 1421–1432
Yeh, F. C. (2011), ‘NTU-90: A high angular resolution brain atlas constructed by q-space diffeomorphic reconstruction’, NeuroImage, 58(1), 91–99