Spatiotemporal patterns of beta desynchronization and gamma synchronization in corticographic data during self-paced movement

Spatiotemporal patterns of beta desynchronization and gamma

synchronization in corticographic data during self-paced movement

G. Pfurtschellera,b,*, B. Graimanna, J.E. Hugginsc, S.P. Levinec,d, L.A. Schuhe

aDepartment of Medical Informatics, Institute of Biomedical Engineering, University of Technology Graz, Inffeldgasse 16a/II, A-8010 Graz, AustriabLudwig Boltzmann Institute for Medical Informatics and Neuroinformatics, Inffeldgasse 16a/II, A-8010 Graz, Austria

cDepartment of Biomedical Engineering, University of Michigan, Ann Arbor, MI, USAdDepartment of Physical Medicine and Rehabilitation, University of Michigan, Ann Arbor, MI, USA

eDepartment of Neurology, Henry Ford Hospital, Detroit, MI 48202, USA

Accepted 14 February 2003

Abstract

Objective: To study the spatiotemporal pattern of event-related desynchronization (ERD) and event-related synchronization (ERS) in

electrocorticographic (ECoG) data with closely spaced electrodes.

Methods: Four patients with epilepsy performed self-paced hand movements. The ERD/ERS was quantified and displayed in the form of

time–frequency maps.

Results: In all subjects, a significant beta ERD with embedded gamma ERS was found.

Conclusions: Self-paced movement is accompanied not only by a relatively widespread mu and beta ERD, but also by a more focused

gamma ERS in the 60–90 Hz frequency band.

q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Electrocorticogram; Event-related desynchronization; Event-related synchronization; Self-paced movement; Beta activity; Gamma activity

1. Introduction

Subdural electrodes are closer to neuronal structures in

superficial cortical layers than electroencephalogram (EEG)

electrodes placed on the scalp. It is estimated by Cooper

et al. (1965) that scalp electrodes represent the spatially

averaged electrical activity over a cortical area of at least

several square centimeters. Several closely spaced subdural

electrodes can be placed over an area of this size such that

each of these electrodes measures the spatially averaged

bioelectrical activity of an area very likely much smaller

than several square centimeters. The advantages of subdural

recordings include recording from smaller sources of

‘synchronized activity’, higher signal-to-noise ratio than

scalp recordings, and increased ability to record and study

gamma activity above 30 Hz. Gamma activity is generated

by rapidly oscillating cell assemblies comprised of a small

number of neurons (Singer, 1993). Consequently, gamma

activity is characterized by small amplitude fluctuations that

are not easily recordable with scalp electrodes. Sheer et al.

(1966) was the first to report movement-induced gamma

activity in scalp EEG. Gamma bursts of 40 Hz embedded in

desynchronized mu and central beta rhythms were recorded

during a finger movement task in an able-bodied subject by

Pfurtscheller et al. (1993).

Self-paced limb movements are accompanied by 3

different types of event-related desynchronization and

synchronization (ERD/ERS) patterns on scalp EEG: (i)

contralateral dominant alpha and beta ERD prior to

movement; (ii) bilateral symmetrical alpha and beta ERD

during execution of movement and (iii) contralateral

dominant beta rebound (beta ERS) within the first second

after movement-offset (Pfurtscheller and Lopes da Silva,

1999a). Additionally, in subdural recordings, contralateral

dominant gamma bursts (gamma ERS) are expected during

the execution phase.

The aim of the present investigation is to study the

dynamics of oscillatory activity in the alpha, beta, and

Clinical Neurophysiology 114 (2003) 1226–1236

www.elsevier.com/locate/clinph

1388-2457/03/$30.00 q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Science Ireland Ltd. All rights reserved.

doi:10.1016/S1388-2457(03)00067-1

* Corresponding author. Tel.: þ43-316-873-5300; fax: þ43-316-873-

5349.

E-mail address: [email protected] (G. Pfurtscheller).

gamma bands during self-paced movement in the form of a

palmar pinch and tongue protrusion. We studied the

spatiotemporal characteristics of ERD/ERS patterns in

electrocorticographic (EcoG) data recorded from 4 patients

who participated in an epilepsy surgery program in the

course of which 62–120 subdural electrodes had been

implanted over frontal and parietal areas.

2. Methods

2.1. Subjects and experimental paradigms

Subjects participating in this study were patients in the

comprehensive epilepsy program at the Henry Ford Hospital

in Detroit, Michigan, who had undergone implantation of

subdural electrodes for the purpose of epilepsy surgery.

Electrode locations were selected exclusively as part of their

epilepsy surgery evaluation and no consideration was made

for research purposes. The 4 subjects, 3 females and one

male, had an average age of 29 ^ 8.7 years. Additional

information about the participating subjects is listed in

Table 1. The subjects performed 2–4 groups of 50

repetitions of a palmar pinch movement and tongue

protrusion in a self-paced manner. The execution of the

requested hand movement was documented by electromyo-

gram (EMG) electrodes placed on the first dorsal inteross-

eous muscle and the ulnar styloid process. The occurrence

of the tongue movement was documented by EMG

electrodes placed above and below opposite corners of the

mouth. ECoG and EMG were recorded on a Nicolet BMSI

5000 at a sampling rate of 200 Hz with bandpass-filter

settings between 0.55 and 100 Hz. More details of the

applied recording technique are described elsewhere

(Levine et al., 2000).

2.2. Quantification of ERD/ERS

The quantification of ERD/ERS was done in 5 steps:

bandpass filtering, extraction of trials, squaring of samples,

averaging over trials, and smoothing of the ERD/ERS

values. The first step, bandpass filtering, was performed in

the frequency domain. The ECoG signal was transformed

into the frequency domain by the fast Fourier transform

(FFT). After multiplying by the bandpass-filter function, the

signal was inverse-transformed to obtain a filtered signal in

the time domain. The bandpass-filtered ECoG was then

segmented into trials of 6 s in length (1200 samples) and

aligned to the onset of the trigger event (600 samples before

and 600 samples after the trigger). Due to the self-paced

recording protocol, the time periods between consecutive

movements could be shorter than 6 s. This implies the

possibility of overlapping consecutive trials and thus a

compromise between the number of trials available for

Table 1

Characteristics of the participating subjects

Subject C17 Age 19 Sex Female

#Electrodes and grid placement 120(32)/left frontal parietal grid

Seizure type Localization-related epilepsy with secondarily generalized tonic–clonic

seizures from both hippocampi

Seizure etiology Cryptogenic; seizure frequency worsened following a closed head injury

Cortical abnormalities, physical impairments None

Subject C18 Age 29 Sex Female

#Electrodes and grid placement 62(8)/sensorimotor strip

Seizure type Localization-related epilepsy with complex partial seizures of left posterior

temporal neocortical onset

Seizure etiology Cryptogenic

Cortical abnormalities, physical impairments Prior left temporal lobectomy

Subject C20 Age 26 Sex Male

#Electrodes and grid placement 114(36)/right parietal grid

Seizure type Localization-related epilepsy with secondarily generalized tonic–clonic

seizures of right parietal-posterior temporal neocortical onset

Seizure etiology Unclear; premature by 4 weeks but no neonatal seizures or developmental

delay; probably cryptogenic

Cortical abnormalities, physical impairments None

Subject C24 Age 40 Sex Female

#Electrodes and grid placement 112(20)/left partietal frontal grid

Seizure type Localization-related epilepsy with complex partial seizures of regional

right fronto-central onset

Seizure etiology Stroke at age 10 months

Cortical abnormalities, physical impairments Left hemiparesis and severe loss of brain volume of the right; hemisphere

Beside the total number of electrodes, the investigated electrodes (in brackets), the location of the implanted electrode grid/strip and other parameters of

interest are indicated.

G. Pfurtscheller et al. / Clinical Neurophysiology 114 (2003) 1226–1236 1227

ERD/ERS analysis and number of overlapping trials was

necessary. Therefore, trials with a trigger-to-trigger interval

shorter than 4.5 s were not considered. This left a minimum

of 30 trials in each group of actions for further analysis.

The ERD/ERS is defined as the proportional power

decrease (ERD) or power increase (ERS) in relation to the

baseline activity determined from a specific reference

interval which is usually placed several seconds before

trigger onset (Pfurtscheller and Lopes da Silva, 1999b). For

this study, the occasional overlap occurring between trials

made it difficult to choose an appropriate placement of the

reference interval. Therefore, the baseline activity was

derived from the whole trial interval, i.e. the whole trial was

taken to be the reference interval. Since evoked potentials

can mask induced activities, the mean of the averaged data

was subtracted from each sample before squaring the data.

In mathematical terms, the procedure can be expressed as

follows:

Aj ¼1

M 2 1

XM

i¼1

ðxij 2 �xjÞ2 ð1Þ

R ¼1

k

Xr0þk

r0

Aj ð2Þ

ERDj ¼Aj 2 R

R£ 100% ð3Þ

where, M is the total number of trials, xij is the j-th sample of

the i-th trial of the bandpass-filtered data, �xj is the mean of

the j-th sample averaged over all bandpass-filtered trials,

and Aj is the j-th power sample. R is the average power in the

reference interval [r0, r0 þ k ] with r0, the first sample of the

trial and r0 þ k, the last sample of the trial.

The statistical significance of the ERD/ERS values was

verified by applying a t-percentile bootstrap to calculate

confidence intervals with a ¼ 0:01. An ERD/ERS value

was considered as significant with a 99% confidence when

both confidence values of this sample show the same sign.

Details about the bootstrapping of ERD/ERS values can be

found elsewhere (Graimann et al., 2002).

By following this procedure, ERD/ERS values of 10 Hz

bands ranging from 5 to 90 Hz and with an overlap of 5 Hz

were calculated. All values for one ECoG channel were

subsequently used to construct time–frequency plots

(ERD/ERS maps). ERD/ERS maps only display significant

power decrease (ERD, colored in red) or increase (ERS,

colored in blue). Together with the topographical infor-

mation, they provide a clear overview of the movement-

related behavior of the non-phase locked activity over a

broad frequency range. Concrete information on frequency

components occurring at specific electrode locations and the

type of significant reactivity is immediately accessible.

3. Results

In all 4 subjects, movement-related desynchronization in

the alpha and beta bands and synchronization in the gamma

band were observed. The gamma ERS included frequency

components from 60 to 90 Hz, was short-lasting (,1 s) and

was always embedded in mu and/or beta desynchronization.

Examples of time–frequency ERD/ERS maps for a left

parietal frontal electrode (subject C24) and corresponding

ERD/ERS time courses for 4 electrodes arranged in one row

are displayed in Figs. 1 and 2.

The first column in Fig. 1 represents 4 electrodes

overlaying the pre-central region, the next two columns

corresponds to the post-central area and the last columns to

the parietal cortex. A broad banded ERD was present on the

majority of electrodes, whereas, gamma activity was

focused to electrodes overlaying pre- and post-central

regions. Examples of time courses calculated for the 4

electrode positions in the fourth row in Fig. 1 are displayed

in Fig. 2.

The time courses in Fig. 2 show a significant power

decrease in the 20–30 Hz band with a duration of more than

1 s and a power increase in the 70–80 Hz band with a

duration of less than 1 s. On the right-most electrode (H44),

neither a significant beta power decrease nor gamma power

increase was present. It is of interest to note that the

electrode with the largest beta power decrease (electrode

H42) showed also the largest gamma power increase.

Another example (subject C17) of time–frequency

ERD/ERS maps is presented in Fig. 3. The electrode grid

is located over the pre-motor, pre-, post-central, and parietal

areas. From the 7 electrodes showing significant gamma

ERS (.100%), one was located over pre-central, two over

post-central, and 4 over parietal areas. The largest gamma

response was on electrode LC44 placed over the parietal

cortex. In addition to the gamma ERS, the majority of

electrodes displayed beta ERS. The post-movement power

increase over the respective electrode positions always

occurred in close relationship to the previous beta power

decrease.

In subject C20, beside the parietal electrode grid, an

electrode strip was placed over the right supplementary

motor area (SMA). In this subject, there was a characteristic

difference between parietal and SMA electrodes. Only the

former displayed a significant alpha power decrease,

whereas a beta power decrease/increase and a gamma

power increase was present in both regions (Fig. 4). Of

interest is the broadband and large beta ERS on electrode

D26 in the SMA electrode strip and the presence of a beta

ERS without a preceding beta ERD on SMA electrode D27

and parietal electrodes I40 and I39 as well.

A comparison between two movement types (palmar

pinch vs. tongue protrusion) is displayed in Fig. 5 (subject

C18). Two electrodes separated by 1 cm and located over

the Rolandic region showed clearly distinguishable ERD/

ERS patterns. From the 6 electrodes indicated in Fig. 5, the

G. Pfurtscheller et al. / Clinical Neurophysiology 114 (2003) 1226–12361228

Fig. 1. Time–frequency ERD/ERS maps for the left parietal frontal grid in subject C24 triggered on palmar pinch onset at 3 s. Red color indicates significant ðP , 0:01Þ power decrease (ERD) and blue color

significant (.0.01) power increase (ERS). The locations of the electrodes are indicated. In addition, two examples of bandpower time courses for beta and gamma bands are displayed in the left lower corner.

G.

Pfu

rtscheller

eta

l./

Clin

ical

Neu

rop

hysio

log

y1

14

(20

03

)1

22

6–

123

61

22

9

Fig. 2. Bandpower time courses calculated in the bands 20–30 Hz and 70–80 Hz for 4 closely spaced ECoG electrodes in subject C24. The mean values and the 95% confidence intervals are displayed in the form

of curves. Significant changes are indicated by black bars above the x-axis. Note the occurrence of maximal gamma power increase during maximal attenuation of power in the beta band.

G.

Pfu

rtscheller

eta

l./

Clin

ical

Neu

rop

hysio

log

y1

14

(20

03

)1

22

6–

12

36

12

30

Fig. 3. Time–frequency ERD/ERS maps from the left frontal parietal grid in subject C17 triggered on palmar pinch onset at 3 s. For further explanation, see Fig. 1.

G.

Pfu

rtscheller

eta

l./

Clin

ical

Neu

rop

hysio

log

y1

14

(20

03

)1

22

6–

123

61

23

1

Fig. 4. Time–frequency ERD/ERS maps from the right parietal grid and a strip over the right supplementary motor area (SMA) in subject C20. For further explanation, see Fig. 1.

G.

Pfu

rtscheller

eta

l./

Clin

ical

Neu

rop

hysio

log

y1

14

(20

03

)1

22

6–

12

36

12

32

Fig. 5. Bandpower time courses calculated in the frequency bands 10–20 Hz and 80–90 Hz on two electrodes spaced 1 cm apart in subject C18. The corresponding electrodes are indicated. The upper panel

displays data corresponding to palmar pinch movements, the lower panel those corresponding to tongue protrusions. For further information please see Fig. 2.

G.

Pfu

rtscheller

eta

l./

Clin

ical

Neu

rop

hysio

log

y1

14

(20

03

)1

22

6–

123

61

23

3

first 4 were overlaying the post-central region and the last

two the parietal cortex. The palmar pinch was accompanied

by a significant power decrease in the 10–20 Hz band,

followed by a power increase after hand movement-offset

on the more posterior electrode (FC23). In addition, there

was also a significant power increase of close to 100% in the

80–90 Hz band. Tongue protrusion was accompanied by a

weak power decrease in the 10–20 Hz band and no

augmentation of gamma power on electrode FC23.

The more anterior electrode (FC22) showed similarly a

significant power decrease/increase in the 10–20 Hz band

but no significant power increase in the gamma band during

hand movement (upper left diagrams in Fig. 5). Tongue

protrusion, however, resulted on this electrode in a

significant power increase in the gamma band embedded

in a power decrease in the 10–20 Hz band.

Summarized results focused on the occurrence of ERD

and ERS in the 4 subjects are depicted in Table 2. Out of the

96 investigated electrodes, 34 showed a beta ERD

(significant power decrease of .50%) and 20 electrodes

showed a gamma ERS (significant power increase of

.100%). A post-movement beta rebound (beta ERS) was

only present on 17 electrodes. This means that the most

frequent decline found was movement-related beta

desynchronization.

The locations corresponding to the electrodes displaying

significant ERD/ERS patterns as indicated in Table 2 are

summarized in Table 3. It has to be mentioned again that the

position of the electrodes were selected exclusively as part

of subject’s epilepsy surgery evaluation and no consider-

ation was made for research purposes. Nevertheless, the

post-central- and parietal dominance of the gamma response

is evident in the data reported.

4. Discussion

The spread of ERD and/or ERS over closely spaced

subdural electrodes mounted in a grid (strip) can be used to

estimate the size of the source of synchronized activity. As

indicated in Table 2, in different subjects, 5–11 electrodes

in close proximity contributed to the beta ERD. This may be

interpreted as evidence that the spatial extent of the source

of the desynchronized beta activity appeared in the range of

at least some square centimeters, whereas, the source of

synchronized gamma activity (3–7 electrodes, see Table 2)

was evidently smaller. The gamma activity is expected to

originate from a much smaller area, since such relatively

high frequency components are associated with small

amplitudes (Singer, 1993). The further argument for a

relatively small source of gamma activity is that an

intermediate-range synchrony of 1–3 cm can be expected

in the case of subdural recordings (Nunez, 1995). So for

example, Bullock et al. (1995) measured the subdural

human coherence with 2 mm diameter electrodes and found

that the coherence falls to zero at electrode separations

greater than about 2 cm.

It should explicitly be noted that the percentage gamma

power increase has to be interpreted cautiously. The time–

frequency maps and the time courses represent percentage

changes of band power. The percentage changes are referred

to the total trial length and, therefore, always depend on the

level of corresponding oscillatory activity in the ongoing

ECoG (Pfurtscheller and Lopes da Silva, 1999b). If the level

of gamma activity in each trial is low and the gamma bursts

are of short duration, a high percentage value of gamma

power is obtained.

By comparing the ERD/ERS maps of all 4 subjects,

different types of reactivity patterns during hand movement

can be detected. Subject C24 (Fig. 1) showed relatively

widespread ERD, hardly any post-movement beta ERS, and

a more focused gamma ERS. It is not clear if the diffuse

right cerebral volume loss in subject C24 could have had an

effect on the location and distribution of the ERS and ERD.

Subject C17 (Fig. 3) displayed relatively widespread beta

ERD and gamma ERS and subject C20 (Fig. 4) displayed

beta ERD, gamma ERS, and especially a prominent post-

movement beta ERS. Taking these observations together, it

can be concluded that the 3 phenomena (beta ERD, post-

movement beta ERS, and gamma ERS) do not exhibit the

Table 2

Number of ECoG electrodes displaying significant ðP , 0:01Þ ERD and/or ERS in the alpha and beta bands and significant ðP , 0:01Þ ERS in the gamma band

Subject #Electrodes on grid/strip ERD .50% (5–15 Hz) ERD .50% (20–30 Hz) ERS .100% (20–30 Hz) ERS .100% (60–90 Hz)

C17 32 2 11 5 7

C18 8 5 5 2 3

C20 36 6 7 9 7

C24 20 11 11 1 3

Sum 96 24 34 17 20

Table 3

Location of ECoG electrodes displaying significant ERD/ERS patterns as

indicated in Table 2

SMA Pre-central Post-central Parietal Sum

C17 1 2 4* 7

C18 2* 1 3

C20 1 4 2* 7

C24 1 2* 3

Sum 1 2 10 7 20

The location of the electrode with the most significant ERD/ERS is

marked by an asterisk.

G. Pfurtscheller et al. / Clinical Neurophysiology 114 (2003) 1226–12361234

same behavior, but rather display some degree of indepen-

dence. Their occurrence may depend not only on the cortical

state prior to movement and on the amount of ongoing

rhythmic components, but also on the fact that different

neuronal networks with different properties are involved in

the dynamic behavior of oscillatory activity in different

frequency bands.

Fig. 5 shows that two movement types (palmar pinch and

tongue protrusion) can be distinguished by two closely

spaced subdural electrodes especially when the gamma ERS

is included in the evaluation. This result is in agreement

with the data reported by Crone et al. (1998a,b) who used a

similar electrode montage and processing technique but a

slightly different experimental paradigm. In Crone’s studies,

the subjects were instructed to perform a sustained

voluntary muscle contraction, as objects were visually

displayed depicting the body part to be moved. Each

voluntary movement lasted 3 s, with different movements

(tongue protrusion, eye-winking, fist-clenching, or dorsi-

flexion of one foot) indicated on a monitor in a random

sequence. The subject had to first observe the pictorial

stimulus, then, decide which type of movement to initiate,

and finally, execute the sustained movement. Crone et al.

(1998b) found a broadbanded gamma augmentation of

frequency components between 60 and 100 Hz in the ECoG

data. This induced gamma activity was highly movement-

specific, whereas, the beta ERD was less movement-

specific, but somatotopically more localized than the alpha

ERD (Crone et al., 1998a).

As shown in Figs. 1 and 3, there is no clear beta ERD

difference on electrodes placed anterior and posterior to the

central sulcus. This is surprising because, there is evidence

that central beta activity is generated in the motor cortex.

Dipole source localization from magnetoencephalogram

(MEG) data revealed that the source of central beta

oscillations is in the primary motor area and that the source

of the mu rhythms is in the somatosensory area (Salmelin

and Hari, 1994). Papakostopolulos et al. (1980) presented

ECoG data that revealed beta activity over pre-central and

alpha band activity over post-central areas, with electrodes

that had a distance of 4 mm. One reason for the smeared

differences in the ERD/ERS patterns between pre- and post-

central areas observed in our ECoG data may be due to the

fact that our subjects have epilepsy and this pathophysio-

logical condition may have had an unknown effect on the

ERD/ERS patterns. However, none of the subjects except

subject C24 had cortical abnormalities under the electrodes

from which ECoG was analyzed. Our results are, however,

comparable to those from a similar study performed by

Crone et al. (1998a) in epileptic patients.

A comparison between scalp EEG and subdural ECoG

reactivity patterns related to self-paced hand/finger move-

ment revealed some interesting results. Scalp recordings

show a clear difference between lower (8–10 Hz) and upper

(10–12 Hz) frequency components, with a more focused

desynchronization pattern of the upper frequency com-

ponents close to the electrodes overlaying the contralateral

hand representation areas (Pfurtscheller and Berghold,

1989; Toro et al., 1994; Pfurtscheller et al., 2000). In

contrast to this, no significant difference between lower and

upper alpha band was observed in our ECoG data. This is

also consistent with the findings in ECoG of Arroyo et al.

(1993), Torro et al. (1994), and Crone et al. (1998a).

Both scalp EEG and subdural ECoG recordings show a

similar spread of the movement-related alpha (mu) ERD to

the parietal lobe. The involvement of the posterior parietal

cortex (PPC) in the generation of mu ERD was for the first

time discussed by Babiloni et al. (1999). Such a parietal

alpha band ERD was present in all investigated subjects (see

Figs. 1, 3, and 4). Also, Crone et al. (1998a) observed an

alpha ERD over the inferior parietal lobe during movements

of all investigated body parts. In addition to the parietal

ERD, we found also a gamma ERS on subdural electrodes

placed over the parietal cortex. In the traditional view, the

PPC is a sensory ‘association’ area, but more recently, it was

reported (Andersen and Buneo, 2002) that broadbanded

gamma oscillations in the local field potentials are found in

PPC during planning of a movement. This implies that

neurons in the PPC are also involved in sensory–motor

transformation and the observed alpha band ERD and

induced gamma oscillations may reflect a part of this

transformation process.

The involvement of the SMA in the preparation of

movement is well documented by functional magnetic

resonance imaging (fMRI) studies (for example, Wildgruber

et al., 1997; Ball et al., 1999). A quite recent finding,

however, is that the dynamics of oscillatory activity in the

beta band originating in the SMA is modified by self-paced

finger movement. Exemplarily, Ohara et al. (2000) observed

a beta ERD in the 18–22 Hz frequency band on ECoG

electrodes in the SMA starting 3.4 ^ 0.5 s (mean ^ SD)

prior to movement-onset. This beta ERD was followed by a

fast recovery and weak beta ERS. In contrast to Ohara’s

data, we observed a beta ERD followed by a broadbanded,

large amplitude beta ERS in ECoG recordings from the

SMA and, what is novel, a short-lasting gamma ERS.

Recently, Pfurtscheller et al. (2003) reported about move-

ment-related beta bursts with frequencies around 26 Hz on

midcentrally placed EEG electrodes overlaying the SMA.

Bursts of movement-related gamma oscillations around

40 Hz were observed in EEG (DeFrance and Sheer, 1988;

Pfurtscheller et al., 1993) and MEG (Salenius et al., 1996).

In all of our subjects, we found broadband gamma

oscillations in a frequency range from 60 to 90 Hz, similar

to those as observed by Crone et al. (1998b) and Arroyo et al.

(1993) in ECoG. The latter reported a movement-related

less prominent relative power increase in the 40–100 Hz

band, while the former reported highly movement-specific

60–100 Hz gamma oscillations. The most interesting result

is that gamma activity in ECoG is broadbanded and covers a

frequency range reaching from about 60 to 90 Hz. A

possible explanation for this observation may be that each

G. Pfurtscheller et al. / Clinical Neurophysiology 114 (2003) 1226–1236 1235

subdural electrode represents the spatial average over a

greater number of cortical modules and neuronal assem-

blies, respectively, while each module may oscillate in a

slightly different frequency band within the 60–90 Hz

range. An alternative explanation is that gamma activity

shows a dynamic oscillatory behavior with varying

frequencies and/or phases. Further research, however,

needs to be devoted to this topic. The observed dominant

60–90 Hz ERS during movement execution may originate

in neuronal assemblies that are involved in active cortical

control of movements in motor and parietal areas. There-

fore, these oscillations may represent ‘binding’ phenomena

in the motor system, adjusting the activity of different

muscle groups according to the movement requirements.

Acknowledgements

This study was supported by the Austrian Fonds zur

Foerderung der wissenschaftlichen Forschung (project

P14831PSY), the National Institute of Health, Bethesda,

Maryland, grant NS 4068108, and the National Institute of

Health (NIH) Bioengineering Research Partnership 1-RO1-

NS40681-01. We thank Mr G. Krausz for producing the

figures.

References

Andersen RA, Buneo CA. Intentional maps in posterior parietal cortex.

Annu Rev Neurosci 2002;25:189–220.

Arroyo S, Lesser RP, Gordon B, Uematsu S, Jackson D, Webber R.

Functional significance of the mu rhythms of human cortex: an

electrophysiologic study with subdural electrodes. Electroenceph clin

Neurophysiol 1993;87:76–87.

Babiloni C, Carducci F, Cincotti F, Rossini PM, Neuper C, Pfurtscheller G,

Babiloni G. Human cortical responses to voluntary finger movements.

Mean movement-related potentials vs. desynchronization of EEG mu

rhythm. Neuroimage 1999;10(6):658–65.

Ball T, Schreiber A, Feige B, Wagner M, Lucking CH, Kristeva-Feige R.

The role of higher-order motor areas in voluntary movement as revealed

by high-resolution EEG and fMRI. Neuroimage 1999;10:682–94.

Bullock TH, McClurne MC, Achimowicz JZ, Iraguri-Madoz VJ, Duckrow

RB, Spencer SS. Temporal fluctuations in coherence of brain waves.

Proc Natl Acad Sci USA 1995;92:11568–72.

Cooper R, Winter AL, Crow HL, Walter WG. Comparison of subcortical,

cortical and scalp activity using electronically indwelling electrodes in

man. Electroenceph clin Neurophysiol 1965;18:217–28.

Crone NE, Miglioretti DL, Gordon B, Sieracki JM, Wilson WT, Uematsu S,

Lesser RP. Functional mapping of human sensorimotor cortex with

electrocorticographic spectral analysis: I Alpha and beta event-related

desynchronization. Brain 1998a;121:2271–99.

Crone NE, Miglioretti DL, Gordon B, Lesser RP. Functional mapping

of human sensorimotor cortex with electrocorticographic spectral

analysis: II. Event-related synchronization in the gamma band. Brain

1998b;121:2301–15.

DeFrance J, Sheer DE. Focused arousal, 40-Hz EEG, and motor

programming. In: Giannitrapani D, Murri L, editors. The EEG of

mental activities. Basel: Karger; 1988. p. 153–68.

Graimann B, Huggins JE, Levine SP, Pfurtscheller G. Visualization of

significant ERD/ERS patterns in multichannel EEG and ECoG data.

Clin Neurophysiol 2002;113:43–7.

Levine SP, Huggins JE, BeMent SL, Kushwaha RK, Schuh LA, Rohde

MM, Passaro EA, Ross DA, Elisevich KV, Smith BJ. A direct brain

interface based on event-related potentials. IEEE Trans Rehabil Eng

2000;8(2):180–5.

Nunez PL. Neocortical dynamics and human EEG rhythms. New York,

NY: Oxford University Press; 1995.

Ohara S, Ikeda A, Kunieda T, Yazawa S, Baba K, Nagamine T, Taki W,

Hashimoto N, Mihara T, Shibasaki H. Movement-related change of

electrocorticographic activity in human supplementary motor area

proper. Brain 2000;123:1203–15.

Papakostopolulos D, Crown HJ, Newton P. Spatiotemporal characteristics

for intrinsic evoked and event related potentials in the human cortex. In:

Pfurtscheller G, Buser P, Lopes da Silva FH, Petsche H, editors.

Rhythmic EEG activities and cortical functioning. Amsterdam: Else-

vier/North-Holland Biomedical; 1980. p. 179–200.

Pfurtscheller G, Berghold A. Patterns of cortical activation during

planning of voluntary movement. Electroencheph clin Neurophysiol

1989;72:250–8.

Pfurtscheller G, Lopes da Silva FH. Event-related desynchronization.

Handbook of electroencephalography and clinical neurophysiology,

Vol. 6. Amsterdam: Elsevier; 1999a. (revised edition).

Pfurtscheller G, Lopes da Silva FH. Event-related EEG/MEG synchroniza-

tion and desynchronization: basic principles. Clin Neurophysiol 1999b;

110:1842–57.

Pfurtscheller G, Neuper C, Kalcher J. 40-Hz oscillation during motor

behavior in man. Neurosci Lett 1993;164:179–82.

Pfurtscheller G, Neuper C, Krausz G. Functional dissociation of lower and

upper frequency mu rhythms in relation to voluntary limb movement.

Clin Neurophysiol 2000;111:1873–9.

Pfurtscheller G, Woertz M, Supp G, Lopes da Silva FH. Early onset of post-

movement beta synchronization in the supplementary motor area

(SMA) during self-paced finger movement in man. Neurosci Lett 2003;

(in press).

Salmelin R, Hari R. Spatiotemporal characteristics of sensorimotor

neuromagnetic rhythms related to thumb movment. Neuroscience

1994;60(2):537–50.

Salenius S, Salmelin R, Neuper C, Pfurtscheller G, Hari R. Human cortical

40 Hz rhythm is closely related to EMG rhythmicity. Neurosci Lett

1996;213:75–8.

Sheer DE, Grandstaff NW, Benignus VA. Behavior and 40-c sec, electrical

activity in the brain. Psychol Rep 1966;19:1333–4.

Singer W. Synchronization of cortical activity and its putative role in

information processing and learning. Annu Rev Physiol 1993;55:

349–74.

Toro C, Deuschl G, Thatcher R, Sato S, Kufta C, Hallett M. Event-related

desynchronization and movement-related cortical potentials on the

ECoG and EEG. Electroenceph clin Neurophysiol 1994;93:380–9.

Wildgruber D, Erb M, Klose U, Grodd W. Sequential activation of

supplementary motor area and primary motor cortex during self-paced

finger movement in human evaluated by functional MRI. Neurosci Lett

1997;227:161–4.

G. Pfurtscheller et al. / Clinical Neurophysiology 114 (2003) 1226–12361236