Magnetoencephalographic cortical rhythms

doi:10.1016/S0167-8760(97)00755-1

International Journal of Psychophysiology

Volume 26, Issues 1–3, June 1997, Pages 51-62

https://doi.org/10.1016/S0167-8760(97)00755-1 Get rights and content

Abstract

We have characterized the magnetic 10- and 20-Hz rhythms recorded with a whole-scalp neuromagnetometer during different conditions. Sources of the posterior 10-Hz (alpha) rhythm clustered mainly around the parieto–occipital sulcus and, to a lesser extent, around the calcarine sulci, with several generators. Temporal Spectral Evolution (TSE) analysis, used to follow event-related changes in the different frequency bands, showed strong dampening of the alpha within 200 ms after the appearance of a visual stimulus and also during visual imagery. Suppression was often followed by a rebound above the baseline level. The rolandic mu rhythm consisted of 10- and 20-Hz components with different reactivity and source locations. The 10-Hz component seems to be mainly somatosensory in origin whereas the 20-Hz signal also receives contributions from the motor cortex, and even shows `motorotopy' in its reactivity: the source locations depend in a somatotopical manner on the site of the moving body part. The frequency composition of the posterior spontaneous activity was disturbed in patients with small infarcts of the medial thalamus. It is shown with simulations that a surprisingly small number of synchronized cortical neurons could generate the major part of the recorded oscillatory signal. Finally, some clarifications are suggested to the terminology of brain rhythms. © 1997 Elsevier Science B.V.

Introduction

In his insightful paper about the alpha rhythm of electroencephalogram, Adrian (1944)conceptually laid the basis for the present-day magnetoencephalographic (MEG) recordings by stating that

"With present methods the skull and the scalp are too much in the way, and we need some new physical method to read through them...

In these days we may look with some confidence to the physicists to produce such an instrument, for it is just the sort of thing they can do..."

E.D. Adrian (1944)Brain Rhythsm, Nature, 153: 360–362.

A neuromagnetometer looks through the skull and the scalp, since these tissues are transparent to the magnetic field and thus do not disturb the MEG signals. Consequently, MEG (for a recent extensive review, see Hämäläinen et al., 1993) provides a non-invasive means to study temporospatial activation patterns in the human cerebral cortex. MEG recordings have been successfully used in studies of stimulus- and task-related cortical signals. Early attempts to localize sources of spontaneous brain rhythms were handicapped by the small coverage of the MEG instruments. Only the emergence of whole-scalp magnetometers has made systematic studies of spontaneous cortical activity feasible. In this paper we briefly discuss some of our recent results in this field.

Section snippets

Source clusters

Fig. 1 shows amplitude spectra of spontaneous activity in one subject. Over the posterior regions the `alpha rhythm' has a major frequency peak at 11.5 Hz and a smaller peak at 22 Hz; this rhythmic signal is clearly dampened by opening of the eyes. Over the rolandic regions, the `mu rhythm', also displaying several frequency peaks (9, 18, and 23 Hz), is not affected by opening/closing of the eyes but diminishes during hand movements.

Magnetic field patterns of the oscillatory signals are

10- and 20-Hz oscillations

The mu-rhythm over the rolandic areas consists of signals at about 10 and 20 Hz, with source areas close to those of the 20-ms response to median nerve stimulation (Tiihonen et al., 1989; Salmelin and Hari, 1994b). The dominance of the mu rhythm in the SI hand area may be related to the large representation of hand, and especially of thumb, in the somatosensory homunculus, reflecting the importance of the hand in human behavior. Practically all subjects show magnetic mu rhythm, at least after

Temporal-spectral-evolution

In the above discussion of signal rebound, we already used information obtained with the `Temporal Spectral Evolution' (TSE) method, which quantifies event-related changes in the spontaneous activity (Salmelin and Hari, 1994b). Fig. 8 shows TSE applied on data from an experiment in which the subject was presented with visual stimuli. To obtain the TSE curves, the signals were first filtered through a passband suggested by spectral analysis and, subsequently, the absolute signal values were

Effect of thalamic infarction

Although it is clear that rhythmic oscillations recorded in scalp EEG as well as in MEG mainly reflect cortical currents, the signals have a close connection to the thalamocortical feedback system. Fig. 11 shows amplitude spectra from the parieto–occipital area of a patient who had a small unilateral infarction of the left tuberothalamic artery. Such patients typically have both memory and speech disorders (Kotila et al., 1994). The spectrum is broadened, resembling diffuse disturbance of

Source clusters and functional significance of cortical rhythms

According to MEG recordings, cortical sensory projection areas have their own rhythms which can be modified separately. The data thus contradict the idea of a global brain rhythm. The observed rhythms, especially those arising from the rolandic cortex, typically consist of 10-Hz basic frequency and of an about 20-Hz component which explains the comb-like shape of the waveforms. The 10- and 20-Hz rhythms seem to have different sources and reactivity and thereby most probably serve different

Acknowledgements

This work was supported by the Academy of Finland and by the Sigrid Jusélius Foundation. We thank P. Hari for useful discussions concerning the synchrony of elements, V. Jousmäki for help in the experiments, and O.V. Lounasmaa and K. Portin for comments on the manuscript. The MR images were obtained at the Department of Radiology, University of Helsinki.

References (30)

Adrian, E.D. (1944) Brain rhythms. Nature, 153:...
Ahonen, A.I., Hämäläinen, M.S., Kajola, M.J., Knuutila, J.E.T., Laine, P.P., Lounasmaa, O.V., Parkkonen, L.T., Simola,...
Basar, E. (1988) EEG-dynamics and evoked potentials in sensory and cognitive processing by the brain. In: E. Basar...
Bente, D. (1964) Die Insuffizienz des Vigilitätstonus. Habilitationsschrift. University of...
Elul, R. (1972) The genesis of the EEG. In: C.C. Pfeiffer and J.R. Smythes (Eds.), International Review of...
Halgren, E., Baudena, P., Heit, G., Clarke, M. and Marinkovic, K. (1994) Spatio-temporal stages in face and word...
Hämäläinen, M., Hari, R., Ilmoniemi, R., Knuutila, J. and Lounasmaa, O.V. (1993) Magnetoencephalography — theory,...
Hari, R. (1990) The neuromagnetic method in the study of the human auditory cortex. In: F. Grandori, M. Hoke and G.L....
Hari, R. (1993) Magnetoencephalography as a tool of clinical neurophysiology. In: E. Niedermeyer and F. Lopes da Silva...
Karhu, J., Hari, R., Paetau, R., Kajola, M. and Mervaala, E. (1994) Cortical reactivity in progressive myoclonus...

Kaufman, L., Schwarz, B., Salustri, C. and Williamson, S. (1990) Modulation of spontaneous brain activity during mental...

Kotila, M., Hokkanen, L., Laaksonen, R. and Valanne, L. (1994) Long-term prognosis after left tuberothalamic...

Llinas, R. (1988) The intrinsic electrophysiological properties of mammalian neurons: insights into central nervous...

Lopes da Silva, F. and Storm van Leeuwen, W. (1977) The cortical source of alpha rhythm. Neurosci. Lett., 6:...

Lu, S.-T., Kajola, M., Joutsiniemi, S.-L., Knuutila, J. and Hari, R. (1992) Generator sites of spontaneous MEG activity...

Cited by (227)

Finger pinching and imagination classification: A fusion of CNN architectures for IoMT-enabled BCI applications
2024, Information Fusion
A Brain–Computer Interface (BCI), integrated with the Internet of Medical Things (IoMT) and based on electroencephalogram (EEG) technology, allows users to control external devices by decoding brainwave patterns. Advanced deep learning-based BCIs, especially those utilizing sensorimotor rhythms (SMRs), have emerged as direct brain–device communication facilitators. SMRs involve users imagining limb motions to induce specific brain activity changes in the motor cortex. Despite progress, some users struggle with BCIs due to weak signals, individual variability, and limited task applicability. This study introduces an unsupervised EEG preprocessing pipeline for SMR-based BCIs. It evaluates an EEG dataset recorded during finger movements, employing two cleaning methods: an investigator-dependent pipeline and our proposed unsupervised method. Two distinct feature datasets are generated: one from cleaned EEG data processed into spectrogram images using supervised preprocessing, and another from data cleaned using our proposed unsupervised pipeline. The study extensively assesses five transfer learning convolutional neural network (TL-CNN) models for distinguishing Motor Imagery (MI) from finger movements (Mex) using the generated datasets. A novel probability fusion technique is developed to enhance TL-CNN classification in Mex versus MI finger-pinching actions. Comparative results show that the fusion-based method outperforms other state-of-the-art methods when applied to unsupervised EEG data. Specifically, our proposed approach achieves 97.9% accuracy, 93.4% precision, 95% recall, and an F1-score of 93.2%, demonstrating significant progress in distinguishing MI and Mex activities through the use of our unsupervised pre-processing pipeline and fusion-based CNN method. Our findings demonstrate the potential of our approach to serve as a benchmark for the global interdisciplinary research community and enable the development of future more effective and user-friendly real-time BCI systems.
Mu rhythm and corticospinal excitability capture two different frames of motor resonance: A TMS–EEG co-registration study
2022, Cortex
Humans are equipped with an extraordinary ability to understand and imitate actions by mapping the observed movement onto their own cortical motor system. Long-established lines of research have identified two correlates of this motor resonance following action observation: the mu rhythm event-related desynchronization (mu-ERD) recorded through electroencephalography (EEG) and the facilitation of motor evoked potentials (MEPs) induced by transcranial magnetic stimulation (TMS) of the primary motor cortex (M1). Yet, whether mu-ERD and MEP facilitation reflect unique or distinct mechanisms is not conclusive, as prior work did not combine simultaneous TMS–EEG recording with a trial-by-trial analysis of the two markers. To address this issue, here, we used TMS–EEG co-registration while participants observed and executed finger movements. EEG was continuously recorded while single-pulse TMS was administered over the left M1 and MEPs were recorded from the right hand. We found stronger motor cortex recruitment during action execution and observation as shown by mu-ERD. MEPs instead were larger overall during action execution and showed a facilitation specific to the muscles involved in the observed movements. Interestingly, when analyzing these two parameters using a trial-by-trial statistical approach, we did not find any relationship between mu-ERD and MEPs within the action observation condition. Our findings support the notion that EEG and TMS indices of motor resonance reflect distinct neural mechanisms.
The natural frequencies of the resting human brain: An MEG-based atlas
2022, NeuroImage
Brain oscillations are considered to play a pivotal role in neural communication. However, detailed information regarding the typical oscillatory patterns of individual brain regions is surprisingly scarce. In this study we applied a multivariate data-driven approach to create an atlas of the natural frequencies of the resting human brain on a voxel-by-voxel basis. We analysed resting-state magnetoencephalography (MEG) data from 128 healthy adult volunteers obtained from the Open MEG Archive (OMEGA). Spectral power was computed in source space in 500 ms steps for 82 frequency bins logarithmically spaced from 1.7 to 99.5 Hz. We then applied k-means clustering to detect characteristic spectral profiles and to eventually identify the natural frequency of each voxel. Our results provided empirical confirmation of the canonical frequency bands and revealed a region-specific organisation of intrinsic oscillatory activity, following both a medial-to-lateral and a posterior-to-anterior gradient of increasing frequency. In particular, medial fronto-temporal regions were characterised by slow rhythms (delta/theta). Posterior regions presented natural frequencies in the alpha band, although with differentiated generators in the precuneus and in sensory-specific cortices (i.e., visual and auditory). Somatomotor regions were distinguished by the mu rhythm, while the lateral prefrontal cortex was characterised by oscillations in the high beta range (>20 Hz). Importantly, the brain map of natural frequencies was highly replicable in two independent subsamples of individuals. To the best of our knowledge, this is the most comprehensive atlas of ongoing oscillatory activity performed to date. Critically, the identification of natural frequencies is a fundamental step towards a better understanding of the functional architecture of the human brain.
Neural correlates of familiar and unfamiliar action in infancy
2022, Journal of Experimental Child Psychology
Behavioral evidence shows that experience with an action shapes action perception. Neural mirroring has been suggested as a mechanism underlying this behavioral phenomenon. Suppression of electroencephalogram (EEG) power in the mu frequency band, an index of motor activation, typically reflects neural mirroring. However, contradictory findings exist regarding the association between mu suppression and motor familiarity in infant EEG studies. In this study, we investigated the neural underpinnings reflecting the role of familiarity in action perception. We measured neural processing of familiar (grasp) and novel (tool-use) actions in 9- and 12-month-old infants. Specifically, we measured infants’ distinct motor/visual activity and explored functional connectivity associated with these processes. Mu suppression was stronger for grasping than for tool use, whereas significant mu and occipital alpha (indexing visual activity) suppression were evident for both actions. Interestingly, selective motor–visual functional connectivity was found during observation of familiar action, a pattern not observed for novel action. Thus, the neural correlates of perception of familiar actions may be best understood in terms of a functional neural network rather than isolated regional activity. Our findings provide novel insights on analytic approaches for identifying motor-specific neural activity while also considering neural networks involved in observing motorically familiar versus unfamiliar actions.
Periodic/Aperiodic parameterization of transient oscillations (PAPTO)–Implications for healthy ageing
2022, NeuroImage
Two techniques for analyzing human extracranial neurophysiological signals, namely the periodic/aperiodic parameterization of neural power spectra and the transient events framework of oscillatory activity, have recently emerged in the scientific literature. In this work, we integrate these two analysis perspectives to analyze extracranial neurophysiological signals as a series of transient rhythmic events disambiguated from the background aperiodic activity. We call this novel technique the periodic/aperiodic parametrization of transient oscillations (PAPTO). We demonstrate PAPTO by investigating resting-state sensorimotor magnetoencephalography recordings from the Cambridge Center for Ageing and Neuroscience cross-sectional study on healthy ageing (n = 600, ages 18–88). We show that PAPTO is more sensitive to neocortical transient beta rhythms compared to more conventional transient event detection algorithms and captures more variance in the resting-state occurrence rate of beta events across participants. The improved sensitivity of PAPTO reveals that the beta occurrence rate almost doubles over the adult lifespan which we discuss in terms of thalamocortical beta generation in the somatosensory cortex and the age-related decline of sensory perception.
Cell type-specific excitability probed by optogenetic stimulation depends on the phase of the alpha oscillation
2022, Brain Stimulation
Alpha oscillations have been proposed to provide phasic inhibition in the brain. Yet, pinging alpha oscillations with transcranial magnetic stimulation (TMS) to examine phase-dependent network excitability has resulted in conflicting findings. At the cellular level, such gating by the alpha oscillation remains poorly understood.
We examine how the excitability of pyramidal cells and presumed fast-spiking inhibitory interneurons depends on the phase of the alpha oscillation.
Optogenetic stimulation pulses were administered at random phases of the alpha oscillation in the posterior parietal cortex (PPC) of two adult ferrets that expressed channelrhodopsin in pyramidal cells. Post-stimulation firing probability was calculated as a function of the stimulation phase of the alpha oscillation for both verum and sham stimulation.
The excitability of pyramidal cells depended on the alpha phase, in anticorrelation with their intrinsic phase preference; pyramidal cells were more responsive to optogenetic stimulation at the alpha phase with intrinsically low firing rates. In contrast, presumed fast-spiking inhibitory interneurons did not show such a phase dependency despite their stronger intrinsic phase preference.
Alpha oscillations gate input to PPC in a phase-dependent manner such that low intrinsic activity was associated with higher responsiveness to input. This finding supports a model of cortical oscillation, in which internal processing and communication are limited to the depolarized half-cycle, whereas the other half-cycle serves as a signal detector for unexpected input. The functional role of different parts of the alpha cycle may vary across the cortex depending on local neuronal firing properties.

View all citing articles on Scopus

View full text