Role of white-matter pathways in coordinating alpha oscillations in resting visual cortex
Introduction
The most salient feature of electrical activity in human neocortex in the absence of explicit cognitive tasks is strong ≈ 10 Hz oscillations (Berger, 1875, Hari and Salmelin, 1997). These posterior α-oscillations are typically recorded over occipital and posterior-parietal regions and are particularly strong within and around the calcarine fissure—where the primary visual cortex (V1) is located—as well as in the occipito-parietal fissure (Hari and Salmelin, 1997, Ciulla et al., 1999). Although initially regarded as functionally irrelevant, evidence is now accumulating that posterior α-oscillations do not merely reflect passive idling of visual areas but correlate with allocation of visuo-spatial attention (Yamagishi et al., 2005, Jensen et al., 2010, Capilla et al., 2012). For example, during anticipatory cue-stimulus intervals, α decreases in those regions of V1 that correspond to attended locations in the visual field and increases in unattended or distractor regions (Kelly et al., 2006, Rihs et al., 2007). Moreover, these modulations do not only pertain to spatial attention tasks but extend to feature-based attentional processes in higher-order visual areas including the dorsal and ventral projection systems (Jokisch and Jensen, 2007, Snyder and Foxe, 2010). Thus, power fluctuations in posterior α-oscillations seem to reflect modulations in cortical excitability, constituting a fundamental mechanism for flexible routing of visual attention (Jensen et al., 2002, Romei et al., 2008, Spaak et al., 2012). Research on the neuronal mechanisms underlying attention-driven α-modulation is expected to benefit from a characterization of the resting-state organization of posterior α-oscillations.
Magnetoencephalographic (MEG) recordings in human subjects and local field potential (LFP) recordings in dogs and macaques have shown that posterior α-oscillations can be recorded throughout the visual system (Lopes Da Silva and Storm van Leeuwen, 1977, Salmelin and Hari, 1994, Hari and Salmelin, 1997, Ciulla et al., 1999, Bollimunta et al., 2008, Bollimunta et al., 2011, Spaak et al., 2012). In addition to cortical sources of α, recordings in behaving dogs and slice preparations of cat lateral geniculate nucleus (LGN) have observed α-sources in thalamic nuclei, particularly the LGN and pulvinar (Lopes da Silva et al., 1973, Hughes et al., 2004). Moreover, the time-courses of sources in LGN and in particular the pulvinar were correlated with various α-sources in occipital cortex (Lopes Da Silva et al., 1980). Furthermore, EEG-fMRI recordings in humans have found resting-state fluctuations in posterior α-power to be correlated with fluctuations in blood-level-oxygenation-level (BOLD) signal throughout the visual system and in several subcortical nuclei (Goldman et al., 2002, Moosmann et al., 2003, Feige et al., 2005). Thus, although posterior α-oscillations seem to involve large-scale thalamo-cortical networks, the nature of their involvement remains controversial (Silva et al., 1991, Karameh et al., 2006).
In particular, it is unclear if α-oscillations are generated at the source locations identified by MEG or if they are generated at other locations and propagate through white-matter pathways. For example, α-oscillations in V1 might be generated within V1 itself (Liley et al., 1999), reflect propagated oscillations that are generated in the LGN (Lopes da Silva et al., 1974, Hughes et al., 2004), which is densely connected to V1 via the optic radiation, or reflect reverberation within thalamo-cortical loops (Robinson et al., 2001, Rennie et al., 2002). Similarly, α-oscillations in different regions of the visual system might be generated locally or reflect propagated oscillations from distant cortical or thalamic regions. In this study, we assessed the contribution of white-matter pathways in the propagation and coordination of posterior α-oscillations. To this end, we combined MEG source-modeling (Woolrich et al., 2011), diffusion tensor imaging (DTI) based probabilistic fiber tracking (Behrens et al., 2003b), and biophysical modeling.
The kind of biophysical model we used in this study is referred to as a neural mass model. Neural mass models have a long tradition (Wilson and Cowan, 1973, Lopes da Silva et al., 1974, Freeman, 2004) and have been applied to several EEG phenomena, including \alpha-oscillations (Lopes da Silva et al., 1974), event-related potentials (Jansen and Rit, 1995), and epileptic seizures (Suffczynski et al., 2004). Neural mass models describe the electrical behavior of a piece of neural tissue in terms of macroscopic quantities and ignore the spatial extendedness of the tissue (Deco et al., 2008). An extension of neural mass models are so-called neural field models which can be thought of as consisting of a sheet of neural masses and describe the electrical behavior of neocortex in a spatially continuous manner (Deco et al., 2008). Neural fields have a long tradition as well (Wilson and Cowan, 1973, Nunez, 1974, Wright and Liley, 1995) and also have been applied to several EEG phenomena including delta, alpha, beta, and gamma oscillations (Nunez et al., 2001, Liley and Cadusch, 2002, Rennie et al., 2000, Rennie et al., 2002, Robinson et al., 2001), sleep (Robinson et al., 2002, Steyn-Ross et al., 2005), and general anesthesia (Bojak and Liley, 2005, Hutt and Longtin, 2010, Hindriks and van Putten, 2012). They provide a theoretical framework in which different EEG phenomena can be integrated and their relationships be investigated (Robinson et al., 2001, Breakspear et al., 2006).
The motivation for using a neural mass model in the present study is that they make more feasible an initial investigation into how posterior \alpha-oscillations might emerge from the topology of white-matter pathways and provide a direction for more extented modeling studies. It is of interest to note though, that the combination of neural mass models with white-matter topological data has proven effective in modeling the emergence of resting-state networks (RSNs) in blood-oxygenation level-depend(BOLD) functional magnetic resonance imaging (fMRI) imaging (Ghosh et al., 2008, Deco et al., 2009, Deco et al., 2011, Deco et al., 2013, Honey et al., 2009, Cabral et al., 2011). Thus, the current study should be regarded as an initial orientation that provides a startingpoint for constructing more extended models of the spatio-temporal behavior of \alpha-oscillations in human cortex.
We found that the assumption of a single α-source in the calcarine sulcus (V1) could explain the source-strengths of α-oscillations throughout the occipital lobe, medial posterior–parietal cortex and temporal lobes. Furthermore, the source-strengths of α-oscillations in these regions correlated with both the functional and anatomical connections to V1, consistent with the assumption of a generator in V1. Although this study does not rule out the possibility that α-oscillations are generated throughout the cortex (Robinson et al., 2001, Rennie et al., 2002, Nunez and Srinivasan, 2006), it establishes a central role of V1-connectivity in coordinating α-oscillations in the visual system at rest.
Section snippets
MEG recordings
Ten subjects (3 males, 20–39 years old, mean 27.9) underwent an eyes-closed resting-state MEG scan lasting 5 min on an Elekta Neuromag (Elekta Neuromag Oy, Helsinki, Finland). Data preprocessing included signal space separation, de-noising with independent component analysis (ICA), source reconstruction and bandpass filtering of the MEG signal. External noise was removed using Signal-Space Separation (SSS) and the data was down-sampled to 200 Hz, using the MaxFilter software (Elekta-Neuromag).
Dynamical workingpoint
To obtain a dynamical workingpoint for the model, we first determined the models' stability boundary in the plane spanned by the cortico-cortical and thalamo-cortical connection-strengths K1 and K2, respectively. In terms of dynamics, restricting the workingpoint to the stable region means that we assume resting-state α-oscillations to emerge from stochastic perturbations of a stable equilibrium state, in agreement with empirical studies (Stam et al., 1999, Hindriks et al., 2011) and in line
Discussion
In this study we combined MEG source-modeling (Woolrich et al., 2011), DTI probabilistic tractography, connectivity-based thalamic segmentation (Behrens et al., 2003a, Behrens et al., 2003b, Behrens et al., 2007), and biophysical modeling to investigate the role of white-matter pathways in coordinating α-oscillations in human cortex at rest. We focused on two scenarios for the generation of α-oscillations, namely, local generation within the Calcarine sulcus (V1) as suggested by MEG
Acknowledgments
The authors thank Ole Jensen from the Donders Institute for Brain, Cognition, and Behaviour, Nijmegen, The Netherlands and Wessel van Wieringen from the VU University Amsterdam, The Netherlands for their valuable conversations regarding this manuscript. In addition, the authors like to thank the reviewers for their constructive comments.
GD was supported by the ERC Advanced Grant: DYSTRUCTURE (n. 295129), by the Spanish Research Project SAF2010-16085 and by the CONSOLIDER-INGENIO 2010 Program
References (87)
- et al.
Probabilistic diffusion tractography with multiple fibre orientations: what can we gain?
Neuroimage
(2007) - et al.
Role of local network oscillations in resting-state functional connectivity
Neuroimage
(2011) - et al.
Dynamic causal modeling of evoked responses in EEG and MEG
Neuroimage
(2006) - et al.
Human cortical oscillations: a neuromagnetic view through the skull
Trends Neurosci.
(1997) - et al.
Meanfield modeling of propofol-induced changes in spontaneous EEG rhythms
Neuroimage
(2012) - et al.
Dynamics underlying spontaneous human alpha oscillations: a data-driven approach
Neuroimage
(2011) - et al.
Intra-cortical propagation of EEG alpha oscillations
Neuroimage
(2014) - et al.
Synchronized oscillations at alpha and theta frequencies in the lateral geniculate nucleus
Neuron
(2004) - et al.
Improved optimization for the robust and accurate linear registration and motion correction of brain images
Neuroimage
(2002) EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis
Brain Res. Brain Res. Rev.
(1999)
Induced alpha band power changes in the human EEG and attention
Neurosci. Lett.
A continuum theory of electro-cortical activity
Neurocomputing
EEG and MEG: relevance to neuroscience
Neuron
The cortical source of the alpha rhythm
Neurosci. Lett.
Relative contributions of intracortial and thalamo-cortical processes in the generation of alpha rhythms, revealed by partial coherence analysis
Electroencephalography
Correlates of alpha rhythm in functional magnetic resonance imaging and near infrared spectroscopy
Neuroimage
A neural mass model of spectral responses in electrophysiology
Neuroimage
The brain wave equation: a model for the EEG
Math. Biosci.
Mechanisms of cortical electrical activity and emergence of gamma rhythm
J. Theor. Biol.
Characterization of spontaneous MEG rhythms in healthy adults
Electroencephalogr. Clin. Neurophysiol.
Layer-specific entrainment of gamma-band neural activity by the alpha rhythm in monkey visual cortex
Curr. Biol.
Dynamics of the human alpha rhythm: evidence for non-linearity?
Clin. Neurophysiol.
Dynamics of non-convulsive epileptic phenomena modeled by a bistable neuronal network
Neuroscience
Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject Brain
Neuroimage
MEG beamforming using Bayesian PCA for adaptive data covariance matrix regularization
Neuroimage
Attentional shifts towards an expected visual target alter the level of alpha-band oscillatory activity in the human calcarine cortex
Cogn. Brain Res.
Alpha oscillations in brain functioning: an integrative theory
Int. J. Psychophysiol.
Combined MEG and EEG source imaging by minimization of mutual information
IEEE Trans. Biomed. Eng.
Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging
Nature
Characterization and propagation of uncertainty in diffusion-weighted MR imaging
Magn. Reson. Med.
Uber das elektrenkephalogramm des menschen
Eur. Arch. Psychiatry Clin. Neurosci.
Modeling the effects of anesthesia on the electroencephalogram
Phys. Rev. E
Connecting mean field models of neural activity to EEG and fMRI data
Brain Topogr.
Towards a model-based integration of co-registered electroencephalography / functional magnetic resonance imaging data with realistic neural population meshes
Philos. Transact. A Math. Phys. Eng. Sci.
Neuronal mechanisms of cortical alpha oscillations in awake-behaving macaques
J. Neurosci.
Neuronal mechanisms and attentional modulation of corticothalamic α oscillations
J. Neurosci.
A unifying explanation of primary generalized seizures through nonlinear brain modeling and bifurcation analysis
Cereb. Cortex
Dissociated α-band modulations in the dorsal and ventral visual pathways in visuospatial attention and perception
Cereb. Cortex
MEG characterization of spontaneous alpha rhythm in the human
Brain
Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space
J. Comput. Assist. Tomogr.
Improved localization of cortical activity by combining EEG and MEG with MRI cortical surface reconstruction: a linear approach
J. Cogn. Neurosci.
Ongoing cortical activity at rest: criticality, multistability, and ghost attractors
J. Neurosci.
The dynamic brain: from spiking neurons to neural the dynamic brain: from spiking neurons to neural masses and cortical fields
PLoS Comput. Biol.
Cited by (35)
Understanding the effect of white matter delays on large scale brain synchrony
2024, Communications in Nonlinear Science and Numerical SimulationRelationship between regional white matter hyperintensities and alpha oscillations in older adults
2022, Neurobiology of AgingCitation Excerpt :However, there are only a few EEG studies that have directly investigated the relationship between alpha oscillations and WMHs or integrity (Babiloni et al., 2011, 2008a; Valdés-Hernández et al., 2010; van Straaten et al., 2012). Previously, local and global disturbances of brain anatomy like white matter microstructure (Hinault et al., 2020; Hindriks et al., 2015; Minami et al., 2020; Valdés-Hernández et al., 2010) have been found to be related to alpha rhythm affecting its peak frequency and power. For instance, a previous study with 222 subjects using the Cuban Human Brain Mapping Project (Valdés-Hernández et al., 2010) provides evidence that alpha peak frequency can be associated with both decrease and increase (depending on the region) in the microstructure of thalamocortical or corticothalamic fibers assessed by Fractional Anisotropy using diffusion tensor imaging.
Age-related differences in network structure and dynamic synchrony of cognitive control
2021, NeuroImageCitation Excerpt :Rather, results show that the decreased structural, clustered organization of white matter architecture leads to changes in functional network efficiency, clustering, and stability over time, resulting in reduced cognitive performance. Findings are consistent with previous computational modeling and animal studies suggesting that oscillatory activity propagates through cortico-cortical white matter pathways (e.g., Hindriks et al., 2015), and that myelin thickness has a dominant influence on conduction velocity (e.g., Fields 2008). Conduction failure resulting from myelin damage has also been associated paralysis, sensory-motor dysfunction, or cognitive impairment (e.g., Fields 2008), and thus appears to disrupt the ability of neural assemblies to synchronize (e.g., Bells et al., 2017; see Suárez et al., 2020, for a review).
Increased brain atrophy and lesion load is associated with stronger lower alpha MEG power in multiple sclerosis patients
2021, NeuroImage: ClinicalCitation Excerpt :Furthermore, Schoonhoven et al reported a decreased alpha peak frequency in cognitively impaired MS patients (Schoonhoven et al., 2019). Interestingly, alpha oscillations at least partially arise from thalamocortical feedback loops (Hindriks et al., 2015) and both thalamic atrophy and alpha oscillations have been consistently linked with cognitive impairment in MS (Houtchens et al., 2007; Razvan et al., 2018; Tewarie et al., 2013) Yet, to the best of our knowledge, we are not aware of studies linking MS-induced structural damage to changes in electrophysiological spectral power as assessed through EEG or magnetoencephalography (MEG).
Posterior brain lesions selectively alter alpha oscillatory activity and predict visual performance in hemianopic patients
2019, CortexCitation Excerpt :Moreover, hemianopic patients show a reduced alpha power in the lesioned hemisphere, resulting in an imbalanced oscillatory alpha activity between the two hemispheres, which suggests an altered interhemispheric functioning in the alpha range. Converging evidence show a central role of low-level visual cortices, such as V1 (Bollimunta et al., 2008), in coordinating and propagating alpha oscillations in the visual system (Hindriks et al., 2015). Specifically, it has been proposed that in resting state, alpha oscillations propagate from lower to higher-order visual areas, providing a default organization of the visual system (Hindriks et al., 2015).