Eventrelated Phenomena Eeg Desynchronization And Synchronization

The electrical activity of the brain is ever changing, depending on both exogenous and endogenous factors. The spatiotemporal patterns of EEG/MEG activity reflect changes in brain functional states. These patterns are apparent as a sequence of maps of scalp activity that change at relatively short time intervals, on the order of seconds or even fractions of seconds. Dietrich Lehmann identified these series of maps as reflecting brain "microstates" and proposed that different modes of mentation are associated with different brain EEG microstates.

A classic phenomenon is the EEG/MEG activity that occurs in response to sensory events, the so-called sensory evoked potentials (EPs). In addition, another class of EEG phenomena of the same kind should be distinguished that consists of those changes of the ongoing EEG activity that precede motor acts, such as the readiness potential, or "bereitschafspotential" of Kornhuber and Deecke, and the premotor potentials. In order to detect EPs or, in more general terms, event-related potentials (ERPs), computer-averaging techniques are commonly used. The basic model underlying this approach is that the evoked activity is time and phase locked to a given event, or stimulus, while the ongoing EEG activity behaves as additive noise. Here, I do not consider this kind of phenomena in detail since this forms a specialized EEG field. Nevertheless, it should be noted that the discovery of evoked activity related to cognitive events has resulted in important contributions to neurocognitive research. One example is the contingent negative variation (CNV), first described by Grey Walter in 1964. The CNV is a slow potential shift with negative polarity at the cortical surface that precedes an expected stimulus and that is related to motivation and attention. Another example is a component of EPs that peaks at approximately 300 msec, with surface positivity, after infrequent but task-relevant stimuli that was discovered by Sutton, Zubin, and John in 1965. This component is called the P300, and it depends more on the meaning of the stimulus and the context of the task than on the physical properties of the stimulus. Still another EP phenomenon with relevant cognitive connotations is the so-called processing negativity described by Naatanen that is a large surface negative wave that can begin as early as 60 msec and can last for 500 msec. It is a sign of selective attention. Neurocog-nitive studies using ERPs and, recently, event-related magnetic fields have been successfully carried out. Such investigations have benefited much from the approach developed by Alan Gevins and collaborators, the so-called EP covariance methodology, that has provided interesting results concerning the cortical processes involved in working memory and in planning of movement. In these studies, the recording of EPs at different brain sites during the sequential processing of cognitive tasks allowed researchers to follow sequential and/or parallel activation of different cortical areas as cognitive tasks evolved. This approach has been particularly successful in studies of brain processes underlying language functions, such as the seminal investigation of Riitta Salmelin and collaborators in Helsinki. Using a whole-head MEG and event-related magnetic fields, they were able to trace the progression of brain activity related to picture naming from Wernicke's area to the parietal-temporal and frontal areas of the cortex. Subsequent MEG studies of the same group in collaboration with that of Pim Levelt of Nijmegen revealed a more refined pattern of the dynamics of cortical activation associated with the successive stages of a psychological model of spoken word generation. These neurocogni-tive processes were approached using the novel methodology of combining fMRI and MEG in order to obtain high-resolution imaging, both in space and in time, of cortical activity during semantic processing of visually presented words. These studies confirmed that in general there is a wave of activity that spreads from the occipital cortex to parietal, temporal, and frontal areas within 185 msec during picture naming. Furthermore, they indicated that the effects of word repetition are widespread and occur only after the initial activation of the cortical network. This provides evidence for the involvement of feedback mechanisms in repetition priming (Fig. 5).

Since the focus of this article is the ongoing EEG/ MEG activity, I consider in more detail a class of EEG/ MEG phenomena that are time locked to an event but

Neuron 62

Eeg Desynchronization

Figure 5 Estimated time courses of MEG signals corresponding to different cortical areas involved in processing of novel and repeated words. The MEG sources were estimated using noise normalization and were fMRI biased toward hemodynamically active cortical areas. The results represent averages across four subjects. Black lines show responses to novel words, and gray lines indicate the responses to repeated words. Waveforms are derived from single cortical locations (each representing 0.5 cm2). Vertically, the z scores are shown: a score of 6 corresponds to a significance level ofp< 10~8 (reproduced with permission from Dale et al., 2000).

Figure 5 Estimated time courses of MEG signals corresponding to different cortical areas involved in processing of novel and repeated words. The MEG sources were estimated using noise normalization and were fMRI biased toward hemodynamically active cortical areas. The results represent averages across four subjects. Black lines show responses to novel words, and gray lines indicate the responses to repeated words. Waveforms are derived from single cortical locations (each representing 0.5 cm2). Vertically, the z scores are shown: a score of 6 corresponds to a significance level ofp< 10~8 (reproduced with permission from Dale et al., 2000).

not phase locked. These phenomena cannot be extracted by simple averaging but may be detected by spectral analysis. An event-related phenomenon may consist of either a decrease or an increase in synchrony of the underlying neuronal populations. The former is called event-related desynchronization (ERD), and the latter is called event-related synchronization (ERS) (Fig. 6). Gert Pfurtscheller and collaborators extensively studied these phenomena. When referring to ERD or ERS, one should specify the corresponding frequency band since, for example, there may be ERD of the alpha band and ERS of the beta band at the same time. The term ERD is only meaningful if the EEG activity during the baseline condition shows a clear spectral peak at the frequency band of interest, indicating the existence of a specific rhythmic activity. Complementarily, the term ERS has meaning only if the event results in the emergence of a rhythmic component, and therewith of a spectral peak that was not detectable under baseline conditions. In general, ERD/ERS reflect changes in the activity of neuronal networks that take place under the influence of specific and/or modulating inputs, which alter the parameters controlling the oscillatory behavior of the neuronal networks.

Even within the alpha frequency band, ERD is not a unitary phenomenon since we have to distinguish at least two patterns of alpha ERD: Lower alpha (7-10 Hz) desynchronization is found in response to almost any kind of task and appears to depend mainly on task complexity. Thus, it is unspecific and tends to be topographically widespread over the scalp. Upper alpha (10-12 Hz) desynchronization has a more restricted topographic distribution, particularly over the parieto-occipital areas, and it is most often elicited by events related to the processing of sensorisemantic information, as shown by the investigations of Wolfgang Klimesch. In general, many psychophysiological variables that cause ERD of rhythms within the alpha frequency range are related to perceptual and memory tasks, on the one hand, and voluntary motor actions, on the other hand. Voluntary movements can result in

Figure 6 Principle of ERD (left) and ERS (right) EEG processing. A decrease of power within a given band (8-11 Hz) indicates ERD and an increase of band power (26-30 Hz) indicates ERS. Note that ERD precedes the trigger, a finger movement, and that ERS follows the trigger (i.e., it occurs at the cessation of the movement) (adapted with permission from Pfurtscheler and Lopes da Silva, 1999).

Figure 6 Principle of ERD (left) and ERS (right) EEG processing. A decrease of power within a given band (8-11 Hz) indicates ERD and an increase of band power (26-30 Hz) indicates ERS. Note that ERD precedes the trigger, a finger movement, and that ERS follows the trigger (i.e., it occurs at the cessation of the movement) (adapted with permission from Pfurtscheler and Lopes da Silva, 1999).

an ERD of the upper alpha (also called mu rhythm) and lower beta bands localized close to the sensorimotor areas. This desynchronization starts about 2 sec prior to movement onset, over the contralateral rolandic region in the case of a unilateral movement, and becomes bilaterally symmetrical immediately before movement execution. It is interesting to note that the ERDs for the different frequency bands have specific topographical distributions, indicating that different cortical populations are involved. For instance, in relation to a hand movement, the 10- to 11Hz mu ERD and the 20- to 24-Hz beta ERD display different maxima over the scalp, although both activities are localized around the central sulcus. The mu rhythm ERD has maximal magnitude more posteriorly than the beta activity, indicating that it is generated mainly in the postrolandic somatosensory cortex, whereas the low beta activity is preferentially generated in the prerolandic motor area. In addition, after a voluntary movement the central region exhibits a localized beta ERS that becomes evident in the first second after cessation of the movement, at a time when the rolandic mu rhythm still presents a desynchronized pattern. The exact frequency of this rebound beta ERS can vary considerably with subject and type of movement. This beta ERS is observed not only after a real movement as shown in Fig. 6 but also after an imagined movement. Furthermore, ERS in the gamma frequency band (approximately around 36-40 Hz) can also be found over the central egions during the execution of a movement, in contrast with the beta ERS that has its maximum after the termination of the movement. A prerequisite for detecting this gamma ERS is that alpha ERD takes place at the same time.

A particular feature of ERD/ERS phenomena that we have recently analyzed is that under some conditions one can find a localized ERD at the same time as ERS in a neighboring region. This antagonistic ERD/ ERS phenomenon can occur between two different modalities—for example, ERD of the alpha over the occipital region elicited by visual stimulation accompanied by ERS of the mu rhythm of the central somatosensory region—but it can also occur within the same modality. For example, a voluntary hand movement can result in an ERD over the cortical area representing the hand and simultaneously an ERS over the cortical area representing the leg/foot (Fig. 7). The opposite can be seen in the case of a voluntary foot movement. We interpret this ERD/ERS antagonistic phenomenon as indicating that at the level of the thalamic reticular nucleus, cross talk between the neuronal networks processing different inputs takes place. This may occur as follows: The specific movement would engage a focal attentional process that results in a desynchronization of a module of thala-mocortical networks called the target module. This attentional signal is most likely mediated by the activation of modulating cholinergic inputs. These cholinergic inputs hyperpolarize the inhibitory neurons of the reticular neurons and depolarize the thalamocortical relay neurons of a given module. Consequently, the thalamocortical feedback loop responsible for the rhythmic activity becomes open, which is reflected in the ERD that is recorded over the corresponding cortical projection areas. At the same time, the neurons of the reticular nucleus that are adjacent to those of the target module become disinhibited. This results in an increase in the gain of the feedback loops to which the latter belong, thus resulting in an increase in the magnitude of the corresponding rhythmic activity that is reflected at the cortex by ERD. In this way, the analysis of ERD/ ERS phenomena has led to the formulation of a hypothesis concerning the neurophysiological processes underlying the psychological phenomenon of focal attention/surround inhibition.

In short, ERD can be interpreted as an electrophy-siolgical correlate of activated cortical areas involved in the processing of sensory, motor, or cognitive information. The mirror image of alpha ERD, of course, is alpha ERS (i.e., a pronounced rhythmic activity within the alpha frequency range), indicating that the corresponding neuronal networks are in a state of reduced activity. Thus, these rhythmic activities are sometimes called ''idling rhythms,'' although one should be cautious about the literal interpretation of this term since the underlying neuronal populations are not really ''idle''— they are always active but may display different dynamical properties. An important point is that ERD and ERS cannot be considered as global properties of the brain. Indeed, ERD and ERS phenomena can be found to coexist in neighboring areas and may affect specific EEG/MEG frequency components differently.

Understanding the significance of ERS of the beta frequency range that typically occurs after a movement has been aided by the observation that at the time that this form of ERS occurs the excitability of the corticospinal pathways decreases, as revealed by transcranial magnetic stimulation. This supports the hypothesis that the postmovement beta ERS corresponds to a deactivated state of the motor cortex. In contrast, the ERS in the gamma frequency band appears to reflect a state of active information

Electroencephalogram Desynchronization

Figure 7 Event-related desynchronization (ERD) and event-related synchronization (ERS) in relation to movement. (a) Average (n = 9) of ERD curves calculated in the alpha and beta bands for a right-hand movement task; recording from C3 (left). The maps were calculated for a 125-msec interval during movement (A) and after movement offset in the recovery period (B). (b) Maps displaying ERD and ERS for an interval of 125 msec during voluntary movement of the hand and movement of the foot. The motor homunculus model is shown on the right-hand side to give an indication of the localization of the hand and the foot cortical areas. (c) (Left) Superimposed ERD curves and beta ERS rebound from eight sessions with right-hand motor imagery in one subject within the frequency band 18-26 Hz and for electrode C3. The average is superimposed. (Right) ERD maps displaying simultaneously occurring ERD (contralateral) and ERS (ipsilateral) during imagery and ERS (contralateral) after motor imagery. Color code: dark areas indicate power decrease (ERD) and light areas power increase (ERS) (reproduced with permission from Pfurtscheller and Lopes da Silva, 1999).

Figure 7 Event-related desynchronization (ERD) and event-related synchronization (ERS) in relation to movement. (a) Average (n = 9) of ERD curves calculated in the alpha and beta bands for a right-hand movement task; recording from C3 (left). The maps were calculated for a 125-msec interval during movement (A) and after movement offset in the recovery period (B). (b) Maps displaying ERD and ERS for an interval of 125 msec during voluntary movement of the hand and movement of the foot. The motor homunculus model is shown on the right-hand side to give an indication of the localization of the hand and the foot cortical areas. (c) (Left) Superimposed ERD curves and beta ERS rebound from eight sessions with right-hand motor imagery in one subject within the frequency band 18-26 Hz and for electrode C3. The average is superimposed. (Right) ERD maps displaying simultaneously occurring ERD (contralateral) and ERS (ipsilateral) during imagery and ERS (contralateral) after motor imagery. Color code: dark areas indicate power decrease (ERD) and light areas power increase (ERS) (reproduced with permission from Pfurtscheller and Lopes da Silva, 1999).

processing. Indeed, recordings by Bressler and collaborators from the monkey motor cortex during the performance of a visual-guided motor task showed increased neuronal activity at relatively high frequen cies that corresponds in time to the gamma ERS found in human. Furthermore, these gamma band activities recorded from the striate and motor cortex were correlated when an appropriate motor response was made but were uncorrelated when no response occurred. Similarly, the group of Pfurtscheller found that there was an increase in coherence between the EEG recorded from the sensorimotor and supplementary motor areas over one hemisphere, within the gamma range, during the performance of contralateral finger movements.

Was this article helpful?

0 0
Anxiety and Depression 101

Anxiety and Depression 101

Everything you ever wanted to know about. We have been discussing depression and anxiety and how different information that is out on the market only seems to target one particular cure for these two common conditions that seem to walk hand in hand.

Get My Free Ebook


Post a comment