Gijs Plomp

YNIMG Journal 2023 Journal Article

The coupling between the spatial and temporal scales of neural processes revealed by a joint time-vertex connectome spectral analysis

• Joan Rué-Queralt
• Valentina Mancini
• Vincent Rochas
• Caren Latrèche
• Peter J. Uhlhaas
• Christoph M. Michel
• Gijs Plomp
• Stephan Eliez

Brain oscillations are produced by the coordinated activity of large groups of neurons and different rhythms are thought to reflect different modes of information processing. These modes, in turn, are known to occur at different spatial scales. Nevertheless, how these rhythms support different spatial modes of information processing at the brain scale is not yet fully understood. Here we use "Joint Time-Vertex Spectral Analysis" to characterize the joint spectral content of brain activity both in time (temporal frequencies) and in space over the connectivity graph (spatial connectome harmonics). This method allows us to characterize the relationship between spatially localized or distributed neural processes on one side and their respective temporal frequency bands in source-reconstructed M/EEG signals. We explore this approach on two different datasets, an auditory steady-state response (ASSR) and a visual grating task. Our results suggest that different information processing mechanisms are carried out at different frequency bands: while spatially distributed activity (which may also be interpreted as integration) specifically occurs at low temporal frequencies (alpha and theta) and low graph spatial frequencies, localized electrical activity (i.e., segregation) is observed at high temporal frequencies (high and low gamma) over restricted high spatial graph frequencies. Crucially, the estimated contribution of the distributed and localized neural activity predicts performance in a behavioral task, demonstrating the neurophysiological relevance of the joint time-vertex spectral representation.

YNIMG Journal 2022 Journal Article

Selective attention involves a feature-specific sequential release from inhibitory gating

• Mattia F. Pagnotta
• David Pascucci
• Gijs Plomp

Selective attention is a fundamental cognitive mechanism that allows our brain to preferentially process relevant sensory information, while filtering out distracting information. Attention is thought to flexibly gate the communication of irrelevant information through top-down alpha-rhythmic (8-12 Hz) functional connections, which influence early visual processing. However, the dynamic effects of top-down influence on downstream visual processing remain unknown. Here, we used electroencephalography to investigate local and network effects of selective attention while subjects attended to distinct features of identical stimuli. We found that attention-related changes in the functional brain network organization emerge shortly after stimulus onset, accompanied by an overall decrease of functional connectivity. Signatures of attentional selection were evident from a sequential release from alpha-band parietal gating in feature-selective areas. The directed connectivity paths and temporal evolution of this release from gating were consistent with the sensory effect of each feature, providing a neural basis for how visual processing quickly prioritizes relevant information in functionally specialized areas.

YNIMG Journal 2021 Journal Article

The connectome spectrum as a canonical basis for a sparse representation of fast brain activity

• Joan Rué-Queralt
• Katharina Glomb
• David Pascucci
• Sébastien Tourbier
• Margherita Carboni
• Serge Vulliémoz
• Gijs Plomp
• Patric Hagmann

The functional organization of neural processes is constrained by the brain's intrinsic structural connectivity, i.e., the connectome. Here, we explore how structural connectivity can improve the representation of brain activity signals and their dynamics. Using a multi-modal imaging dataset (electroencephalography, structural MRI, and diffusion MRI), we represent electrical brain activity at the cortical surface as a time-varying composition of harmonic modes of structural connectivity. These harmonic modes are known as connectome harmonics. Here we describe brain activity signal as a time-varying combination of connectome harmonics. We term this description as the connectome spectrum of the signal. We found that: first, the brain activity signal is represented more compactly by the connectome spectrum than by the traditional area-based representation; second, the connectome spectrum characterizes fast brain dynamics in terms of signal broadcasting profile, revealing different temporal regimes of integration and segregation that are consistent across participants. And last, the connectome spectrum characterizes fast brain dynamics with fewer degrees of freedom than area-based signal representations. Specifically, we show that a smaller number of dimensions capture the differences between low-level and high-level visual processing in the connectome spectrum. Also, we demonstrate that connectome harmonics capture more sensitively the topological properties of brain activity. In summary, this work provides statistical, functional, and topological evidence indicating that the description of brain activity in terms of structural connectivity fosters a more comprehensive understanding of large-scale dynamic neural functioning.

YNIMG Journal 2020 Journal Article

Connectome spectral analysis to track EEG task dynamics on a subsecond scale

• Katharina Glomb
• Joan Rué Queralt
• David Pascucci
• Michaël Defferrard
• Sébastien Tourbier
• Margherita Carboni
• Maria Rubega
• Serge Vulliémoz

We present an approach for tracking fast spatiotemporal cortical dynamics in which we combine white matter connectivity data with source-projected electroencephalographic (EEG) data. We employ the mathematical framework of graph signal processing in order to derive the Fourier modes of the brain structural connectivity graph, or "network harmonics". These network harmonics are naturally ordered by smoothness. Smoothness in this context can be understood as the amount of variation along the cortex, leading to a multi-scale representation of brain connectivity. We demonstrate that network harmonics provide a sparse representation of the EEG signal, where, at certain times, the smoothest 15 network harmonics capture 90% of the signal power. This suggests that network harmonics are functionally meaningful, which we demonstrate by using them as a basis for the functional EEG data recorded from a face detection task. There, only 13 network harmonics are sufficient to track the large-scale cortical activity during the processing of the stimuli with a 50 ​ms resolution, reproducing well-known activity in the fusiform face area as well as revealing co-activation patterns in somatosensory/motor and frontal cortices that an unconstrained ROI-by-ROI analysis fails to capture. The proposed approach is simple and fast, provides a means of integration of multimodal datasets, and is tied to a theoretical framework in mathematics and physics. Thus, network harmonics point towards promising research directions both theoretically - for example in exploring the relationship between structure and function in the brain - and practically - for example for network tracking in different tasks and groups of individuals, such as patients.

YNIMG Journal 2020 Journal Article

Nested oscillations and brain connectivity during sequential stages of feature-based attention

• Mattia F. Pagnotta
• David Pascucci
• Gijs Plomp

Brain mechanisms of visual selective attention involve both local and network-level activity changes at specific oscillatory rhythms, but their interplay remains poorly explored. Here, we investigate anticipatory and reactive effects of feature-based attention using separate fMRI and EEG recordings, while participants attended to one of two spatially overlapping visual features (motion and orientation). We focused on EEG source analysis of local neuronal rhythms and nested oscillations and on graph analysis of connectivity changes in a network of fMRI-defined regions of interest, and characterized a cascade of attentional effects at multiple spatial scales. We discuss how the results may reconcile several theories of selective attention, by showing how β rhythms support anticipatory information routing through increased network efficiency, while reactive α-band desynchronization patterns and increased α-γ coupling in task-specific sensory areas mediate stimulus-evoked processing of task-relevant signals.

YNICL Journal 2018 Journal Article

Automated diagnosis of temporal lobe epilepsy in the absence of interictal spikes

• Thibault Verhoeven
• Ana Coito
• Gijs Plomp
• Aljoscha Thomschewski
• Francesca Pittau
• Eugen Trinka
• Roland Wiest
• Karl Schaller

Objective: To diagnose and lateralise temporal lobe epilepsy (TLE) by building a classification system that uses directed functional connectivity patterns estimated during EEG periods without visible pathological activity. Methods: Resting-state high-density EEG recording data from 20 left TLE patients, 20 right TLE patients and 35 healthy controls was used. Epochs without interictal spikes were selected. The cortical source activity was obtained for 82 regions of interest and whole-brain directed functional connectivity was estimated in the theta, alpha and beta frequency bands. These connectivity values were then used to build a classification system based on two two-class Random Forests classifiers: TLE vs healthy controls and left vs right TLE. Feature selection and classifier training were done in a leave-one-out procedure to compute the mean classification accuracy. Results: The diagnosis and lateralization classifiers achieved a high accuracy (90.7% and 90.0% respectively), sensitivity (95.0% and 90.0% respectively) and specificity (85.7% and 90.0% respectively). The most important features for diagnosis were the outflows from left and right medial temporal lobe, and for lateralization the right anterior cingulate cortex. The interaction between features was important to achieve correct classification. Significance: This is the first study to automatically diagnose and lateralise TLE based on EEG. The high accuracy achieved demonstrates the potential of directed functional connectivity estimated from EEG periods without visible pathological activity for helping in the diagnosis and lateralization of TLE.

YNIMG Journal 2018 Journal Article

Benchmarking nonparametric Granger causality: Robustness against downsampling and influence of spectral decomposition parameters

• Mattia F. Pagnotta
• Mukesh Dhamala
• Gijs Plomp

Brain function arises from networks of distributed brain areas whose directed interactions vary at subsecond time scales. To investigate such interactions, functional directed connectivity methods based on nonparametric spectral factorization are promising tools, because they can be straightforwardly extended to the nonstationary case using wavelet transforms or multitapers on sliding time window, and allow estimating time-varying spectral measures of Granger–Geweke causality (GGC) from multivariate data. Here we systematically assess the performance of various nonparametric GGC methods in real EEG data recorded over rat cortex during unilateral whisker stimulations, where somatosensory evoked potentials (SEPs) propagate over known areas at known latencies and therefore allow defining fixed criteria to measure the performance of time-varying directed connectivity measures. In doing so, we provide a comprehensive benchmark evaluation of the spectral decomposition parameters that might influence the performance of wavelet and multitaper approaches. Our results show that, under the majority of parameter settings, nonparametric methods can correctly identify the contralateral primary sensory cortex (cS1) as the principal driver of the cortical network. Furthermore, we observe that, when properly optimized, the approach based on Morlet wavelet provided the best detection of the preferential functional targets of cS1; while, the best temporal characterization of whisker-evoked interactions was obtained with a sliding-window multitaper. In addition, we find that nonparametric methods provide GGC estimates that are robust against signal downsampling. Taken together our results provide a range of plausible application values for the spectral decomposition parameters of nonparametric methods, and show that they are well suited to characterize time-varying directed causal influences between neural systems with good temporal resolution.

YNIMG Journal 2014 Journal Article

Neural correlates of visual crowding

• Vitaly Chicherov
• Gijs Plomp
• Michael H. Herzog

In visual crowding, target discrimination strongly deteriorates when flanking elements are added. We have recently shown that crowding cannot be explained by simple low-level interactions and that grouping is a key component instead. We presented a vernier flanked by arrays of vertical lines. When the flankers had the same lengths as the vernier, offset discrimination was strongly impaired. When longer flankers were presented, crowding was weaker. We proposed that crowding is strong when the flankers group with the target (equal length flankers). When the target segregates from the flankers, crowding is weaker (long flankers). To understand the neurophysiological mechanisms of grouping in crowding, here, we adapted the above vernier paradigm to a high-density EEG study. The P1 component reflected basic stimulus characteristics (flanker length) but not crowding. Crowding emerged slowly and manifested as a suppression of the N1 component (after 180ms). Using inverse solutions, we found that the N1 suppression was caused by reduced neural activity in high-level visual areas such as the lateral occipital cortex. Our results suggest that crowding occurs when elements are grouped into wholes, a process reflected by the N1 component.

YNIMG Journal 2014 Journal Article

The physiological plausibility of time-varying Granger-causal modeling: Normalization and weighting by spectral power

• Gijs Plomp
• Charles Quairiaux
• Christoph M. Michel
• Laura Astolfi

Time-varying connectivity methods are increasingly used to study directed interactions between brain regions from electrophysiological signals. These methods often show good results in simulated data but it is unclear to what extent connectivity results obtained from real data are physiologically plausible. Here we introduce a benchmark approach using multichannel somatosensory evoked potentials (SEPs) measured across rat cortex, where the structural and functional connectivity is relatively simple and well-understood. Rat SEPs to whisker stimulation are exclusively initiated by contralateral primary sensory cortex (S1), at known latencies, and with activity spread from S1 to specific cortical regions. This allows for a comparison of time-varying connectivity measures according to fixed criteria. We thus evaluated the performance of time-varying Partial Directed Coherence (PDC) and the Directed Transfer Function (DTF), comparing row- and column-wise normalization and the effect of weighting by the power spectral density (PSD). The benchmark approach revealed clear differences between methods in terms of physiological plausibility, effect size and temporal resolution. The results provide a validation of time-varying directed connectivity methods in an animal model and suggest a driving role for ipsilateral S1 in the later part of the SEP. The benchmark SEP dataset is made freely available.

YNIMG Journal 2011 Journal Article

Corrigendum to “Electrical source dynamics in three functional localizer paradigms” [NeuroImage 53 (2010) 257–267]

• Gijs Plomp
• Christoph M. Michel
• Michael H. Herzog

YNIMG Journal 2011 Journal Article

Eye fixation-related potentials in free viewing identify encoding failures in change detection

• Andrey R. Nikolaev
• Chie Nakatani
• Gijs Plomp
• Peter Jurica
• Cees van Leeuwen

We considered the hypothesis that spontaneous dissociation between the direction of attention and eye movement causes encoding failure in change detection. We tested this hypothesis by analyzing eye fixation-related potentials (EFRP) at the encoding stage of a change blindness task; when participants freely inspect a scene containing an unmarked target region, in which a change will occur in a subsequent presentation. We measured EFRP amplitude prior to the execution of a saccade, depending on its starting or landing position relative to the target region. For those landings inside the target region, we found a difference in EFRP between correct detection and failure. Overall, correspondence between EFRP amplitude and the size of the saccade predicted successful detection of change; lack of correspondence was followed by change blindness. By contrast, saccade sizes and fixation durations around the target region were unrelated to subsequent change detection. Since correspondence between EFRP and eye movement indicates that overt attention was given to the target region, we concluded that overt attention is needed for successful encoding and that dissociation between eye movement and attention leads to change blindness.

YNIMG Journal 2010 Journal Article

Electrical source dynamics in three functional localizer paradigms

• Gijs Plomp
• Christoph M. Michel
• Michael H. Herzog

The visual cortex exhibits functional specialization that can be routinely demonstrated using hemodynamic measures like fMRI and PET. To understand the dynamic nature of cortical processes, however, source imaging with a high temporal resolution is necessary. Here, we asked how well distributed EEG source localization (LAURA) identifies functionally specialized visual processes. We tested three stimulus paradigms commonly used in fMRI with the aim to localize striate cortex, motion-sensitive areas, and face-sensitive areas. EEG source localization showed initial activations in striate and extra-striate areas at around 70ms after stimulus onset. These were quickly followed by extensive cortical, as well as subcortical activation. Functional motion and face-selective areas were localized with margins of below 2cm, at around 170 and 150ms, respectively. The results furthermore show for the first time that the C1 component has generators in the insula and frontal eye fields, but also in subcortical areas like the parahippocampus and the thalamus.

YNIMG Journal 2009 Journal Article

Non-retinotopic feature integration decreases response-locked brain activity as revealed by electrical neuroimaging

• Gijs Plomp
• Manuel R. Mercier
• Thomas U. Otto
• Olaf Blanke
• Michael H. Herzog

When presented with dynamic scenes, the brain integrates visual elements across space and time. Such non-retinotopic processing has been intensively studied from a psychophysical point of view, but little is known about the underlying neural processes. Here we used high-density EEG to reveal neural correlates of non-retinotopic feature integration. In an offset-discrimination task we presented sequences of lines for which feature integration depended on a small, endogenous shift of attention. Attention effects were observed in the stimulus-locked evoked potentials but non-retinotopic feature integration was reflected in voltage topographies time-locked to the behavioral response, lasting for about 400 ms. Statistical parametric mapping of estimated current densities revealed that this integration reduced electrical activity in an extensive network of brain areas, with the effects progressing from high-level visual, via frontal, to central ones. The results suggest that endogenously timed neural processes, rather than bottom-up ones, underlie non-retinotopic feature integration.