- Email: [email protected]
Involvement of Primary Motor Cortex in Motor Imagery: A Neuromagnetic Study
6, 201–208 (1997) NI970286
NEUROIMAGE ARTICLE NO.
Involvement of Primary Motor Cortex in Motor Imagery: A Neuromagnetic Study Alfons Schnitzler,1 Stephan Salenius, Riitta Salmelin, Veikko Jousma¨ki, and Riitta Hari Brain Research Unit, Low Temperature Laboratory, Helsinki University of Technology, P.O. Box 2200, FIN-02015 HUT, Helsinki, Finland Received February 12, 1997
Functional brain imaging studies have indicated that several cortical and subcortical areas active during actual motor performance are also active during imagination or mental rehearsal of movements. Recent evidence shows that the primary motor cortex may also be involved in motor imagery. Using wholescalp magnetoencephalography, we monitored spontaneous and evoked activity of the somatomotor cortex after right median nerve stimuli in seven healthy right-handed subjects while they kinesthetically imagined or actually executed continuous finger movements. Manipulatory finger movements abolished the poststimulus 20-Hz activity of the motor cortex and markedly affected the somatosensory evoked response. Imagination of manipulatory finger movements attenuated the 20-Hz activity by 27% with respect to the rest level but had no effect on the somatosensory response. Slight constant stretching of the fingers suppressed the 20-Hz activity less than motor imagery. The smallest possible, kinesthetically just perceivable finger movements resulted in slightly stronger attenuation of 20-Hz activity than motor imagery did. The effects were observed in both hemispheres but predominantly contralateral to the performing hand. The attempt to execute manipulatory finger movements under experimentally induced ischemia causing paralysis of the hand also strongly suppressed 20-Hz activity but did not affect the somatosensory evoked response. The results indicate that the primary motor cortex is involved in motor imagery. Both imaginative and executive motor tasks appear to utilize the cortical circuitry generating the somatomotor 20-Hz signal. r 1997 Academic Press
INTRODUCTION Motor imagery can be defined as conscious mental rehearsal of a motor act without performing any overt
1 Permanent address: Department of Neurology, Heinrich-HeineUniversity, Moorenstrasse 5, 40225 Du¨sseldorf, Germany.
movement and implies that the subject feels himself executing a given action. Motor imagery thus involves mostly a kinesthetic representation of the action (Jeannerod, 1994). Motor imagery improves motor performance and skill acquisition. Consequently, mental rehearsal of movements is a common technique used by athletes to improve their performance. The beneficial influence of motor imagery on motor performance has previously been reassessed and confirmed in several experiments (Feltz and Landers, 1983), although the neurophysiological mechanisms are not well understood. Interestingly, imagined movements seem to obey the same temporal characteristics and physiological and pathophysiological constraints as executed movements (Sirigu et al., 1996), implying that motor imagery and execution use the same neural representations (for a recent review, see Crammond et al., 1997). Regional cerebral blood flow (rCBF) has been found to increase mainly in the supplementary motor area during imagination of sequential finger movements (Roland et al., 1980). More recent positron emission tomography (PET) (Decety et al., 1994; Stephan et al., 1995; Parsons et al., 1995; Grafton et al., 1996) and functional magnetic resonance imaging (fMRI) studies (Rao et al., 1993) revealed activation of a number of cortical and subcortical areas, including the lateral and medial premotor cortex, anterior cingulate areas, ventral opercular premotor areas, superior and inferior parietal cortex, and posterior cerebellar cortex, but not the primary sensorimotor cortex. In contrast to these findings, the most recent fMRI studies have reported involvement of primary motor cortex during imagined movements in normal subjects (Leonardo et al., 1995; Roth et al., 1996; Porro et al., 1996) and also in a subject with a phantom limb (Ersland et al., 1996). Transcranial magnetic stimulation studies of the motor cortex have pointed in the same direction in showing an increase of motor responses during mental simulation of movements (Gandevia and Rothwell, 1987; Hallett et al., 1994; Abbruzzese et al., 1996); however, such effects could also be explained by excitability changes at the spinal level.
201
1053-8119/97 $25.00 Copyright r 1997 by Academic Press All rights of reproduction in any form reserved.
202
SCHNITZLER ET AL.
In the present study, we assessed cortical activity during imagination and execution of manipulatory finger movements with noninvasive whole-scalp magnetoencephalography. The aim of the study was to find out whether or not the primary somatosensory and motor areas are involved in movement imagination. Experiments were designed to simultaneously monitor evoked activity from the primary somatosensory hand projection cortex (SI) and 20-Hz rhythmic activity from the primary motor cortex (Jasper and Penfield, 1949; Salmelin and Hari, 1994; Salmelin et al., 1995; Conway et al., 1995; Salenius et al., 1997b). We observed a differentially graded suppression of the poststimulus 20-Hz rhythmic activity during execution and imagination of manipulating finger movements, while evoked activity in SI was markedly modulated only by movement execution. Our finding provides evidence that motor imagery is accompanied by changes of neuronal activity within the primary motor cortex and that the same neuronal populations are—to a quantitatively different degree—involved in movement imagination and execution. Results have been presented in abstract form (Schnitzler et al., 1995a).
Imagination, subjects were reminded to imagine and not to perform the task. In one subject the effects of motor imagery on evoked and rhythmic cortical activity were compared with experimentally induced deafferentation and paralysis of the right forearm and hand. This condition bears some analogy to motor imagery: action is represented, but not executed, even though the level at which motor output is blocked is different in the two conditions. Also, since tonic afferent muscular spindle discharges due to increased gamma motoneuron activity during motor imagery could account for improvement of performance (Jeannerod, 1995), deafferentation provides an important reference condition to determine the role of spindle afferent activity during motor imagery. Deafferentation was produced by tourniquet ischemia, increasing the pressure of a blood pressure cuff around the right proximal forearm to 220 mm Hg. The right MN was stimulated at the upper arm by 0.3-ms constant current pulses once every second, with the current strength adjusted before ischemia to produce a slight twitch of the wrist flexors. Neuromagnetic Recording
MATERIALS AND METHODS Subjects and Tasks Seven healthy right-handed members of the laboratory personnel (ages 22–34 years; four females, three males) gave their informed consent to participate in the study. The subject was seated comfortably, with eyes closed, while the right median nerve (MN) was stimulated at the wrist with 0.3-ms constant current pulses once every second. The current strength (4 to 7 mA) was adjusted to produce a slight twitch of the thumb. During stimulation, subjects were instructed to perform eight motor tasks in subsequent runs, each lasting about 5 min. During the Rest condition, the subject sat relaxed without any additional task. In the Manipulation condition, subjects performed continuous finger movements manipulating a small cube between thumb and fingers of one hand. In the Imagination condition, subjects were asked to kinesthetically imagine themselves performing the same manipulating movements without actually executing them. In the Stretch condition, subjects were instructed to keep all finger joints slightly stretched to just overcome the passive flexion force. In the Small motion condition subjects had to perform the smallest possible, just subjectively perceivable, extension and flexion movements of the fingers. All tasks were performed separately with each hand. During all tasks, finger movements were monitored visually using a video camera and electrophysiologically with continuous surface electromyography (EMG) of the finger extensor muscles. If any muscle activity different from Rest was noted on the EMG during
Magnetic brain signals were recorded noninvasively with a Neuromag-122 whole-scalp neuromagnetometer in a magnetically shielded room, while the subject rested his or her head against the helmet-shaped bottom surface of the device (Ahonen et al., 1993). Neuromag-122 contains 122 superconducting planar gradiometers which detect the largest signal above an active cortical area. The signals were recorded with a passband of 0.03–330 Hz, digitized at 1 kHz, and stored for off-line analysis. To determine the position of the head with respect to the sensor array, small currents were led to three indicator coils attached to the scalp to produce magnetic signals. The coil positions were related to the head coordinates, defined by the preauricular points and the nasion, with a 3-D digitizer. Data Analysis Two different analysis approaches were employed to the same data set to simultaneously yield information about the task-dependent modulation of cortical evoked activity and spontaneous rhythms. 1. Somatosensory evoked fields (SEFs). Magnetic signals were digitally low-pass filtered at 140 Hz and about 150 single responses were averaged, time locked to the MN stimuli. A time-varying single-dipole model (Ha¨ma¨la¨inen et al., 1993) was employed to explain the measured signals during the first 60 ms after the stimuli. Superposition of the identified dipoles on individual MRI scans confirmed that these sources reflected activity within the contralateral primary somatosensory cortex, SI (Hari and Kaukoranta, 1985; Wood et
MOTOR IMAGERY AND SENSORIMOTOR CORTEX
al., 1985; Allison et al., 1991; Baumgartner et al., 1993; Schnitzler et al., 1995b). The strongest activity of the SI cortex was compared between tasks. 2. Sensorimotor rhythms. Spectral analysis showed prominent rhythmic activity around 10 and 20 Hz. The temporal spectral evolution method (TSE; Salmelin and Hari, 1994) was employed to calculate the average amplitude level of 20-Hz activity as a function of time, with respect to MN stimuli: the signals were filtered through passbands around 20 Hz, as suggested by spectral analysis. The absolute values were then averaged over about 60 trials and smoothed with a 15-Hz low-pass filter. Somatomotor 20-Hz rhythmic activity shows a characteristic bilateral enhancement (‘‘rebound’’) around 400–700 ms after MN stimuli. This rebound is markedly diminished or even abolished during finger movements (Salenius et al., 1997a). To quantify the modulation of the 20-Hz signals in different tasks, we calculated the vector sum of the TSE signals from the most reactive sensor pair above the sensorimotor hand area (measuring the orthogonal derivatives of the magnetic field at that location). Peak TSE amplitudes in the various tasks were compared with two-tailed t tests for paired differences. Sources of the 20-Hz activity were identified from bandpass-filtered data in 50 poststimulus epochs, 300– 1000 ms after the MN stimulus. Dipolar field patterns were sought at 5-ms intervals, and the strengths, orientations, and three-dimensional locations of the corresponding equivalent current dipoles were determined by means of a least-squares fit to the signals measured by a subset of sensors centered over the hand area. Sources were accepted when they accounted for at least 90% of the measured field variance. The sources can be localized with an accuracy of about 5 mm. RESULTS Figure 1 shows representative 10-s traces of the 20-Hz activity (filtered through 18–25 Hz) above the left sensorimotor cortex of Subject 1 during rest (top) and during the four tasks performed with the right hand. In all conditions the right MN was stimulated electrically once every second. In the rest condition, the stimuli were typically followed by a transient burst-like increase of the 20-Hz rhythmic activity, as described previously in detail (Salenius et al., 1997a). During movement imagination the 20-Hz bursts were dampened, and during actual performance of finger movements (Manipulation and Small motion) they were almost completely suppressed. The reduction of the 20-Hz activity during Imagination varied from intervals of complete suppression to periods of rather small or no effect. Figure 2 illustrates the effects of all tasks on the poststimulus 20-Hz burst, quantified by the TSE
203
FIG. 1. 10-s intervals of spontaneous neuromagnetic activity of Subject 1, recorded above the left sensorimotor hand area during the different tasks (Rest, Imagination, Manipulation, Stretch, Small motion). The signals were bandpass filtered through 18–25 Hz. Dashed vertical lines indicate right median nerve stimuli delivered once per second. Note the bursts following each MN stimulus and the suppression of these bursts during the tasks.
method, and on the SEFs to right MN stimulation, recorded from the left sensorimotor cortex. In the rest condition, the rebound of the 20-Hz activity started about 200 ms after stimulus onset and peaked at around 450 ms. Overt manipulatory and smallest possible finger movements abolished the 20-Hz rebound in this subject almost completely. Movement imagination led to a 50% reduction, and simple static finger stretch led to only a slight decrease of the rebound. These effects were most prominent when the task was performed with the stimulated (right) hand. SEFs of the contralateral SI cortex were markedly affected by manipulatory movements (Manipulation) which caused activation of cutaneous and deep afferents. No discernible changes were observed in SI responses during the other conditions (Imagination, Stretch, and Small Motion). Electromyographic activity of finger extensors was largest during Manipulation and only marginal
204
SCHNITZLER ET AL.
FIG. 2. Task-related effects on somatomotor 20-Hz activity, quantified with the TSE method, and EMG and SEFs recorded above the left sensorimotor hand cortex in Subject 1, with the EMG activity plotted below. The different tasks (Rest, Imagination, Manipulation, Stretch, Small motion) are illustrated by different line styles. The left and right columns show the results for the tasks performed with the right and left hands, respectively. (Top row) Temporal evolution of 20-Hz activity. Sixty epochs (each 1 s in duration) were averaged with respect to right MN stimuli. (Middle row) Rectified and averaged EMG of right finger extensors recorded simultaneously with the cortical signals. (Bottom row) SEFs to stimulation of the right MN.
during Imagination. However, there was no direct relation between reduction of the 20-Hz rebound and muscular activity: although EMG activity was higher for Stretch than for Imagination, the 20-Hz activity was more strongly suppressed during Imagination. Figure 3 illustrates the mean (1SEM) effects of the different tasks on the 20-Hz rebound compared with the rest condition. The poststimulus 20-Hz rebound was reduced statistically significantly during all tasks (P , 0.005). The most pronounced suppression was observed during actual movements of the right hand (Manipulation 45% and Small motion 39%), with the
FIG. 3. Mean (1 SEM), seven subjects) amplitudes of the 20-Hz rebound for the different tasks relative to the rest condition (5100%).
smallest effect occurring during simple static muscular activity (Stretch 18%) and an intermediate effect (27%) during Imagination. The 20-Hz activity was also dampened, but to a lesser degree (11–37%; P , 0.01), when the tasks were performed with the left hand. Figure 4 illustrates the current dipole cluster determined from the 20-Hz bursts during Rest in Subject 1, superimposed on the surface image of his brain. During all tasks, the sources covered the same area in the hand motor cortex. Source density plot along the anterior– posterior axis shows that the 20-Hz activity was generated in the motor cortex. Similar results were obtained in the other subjects. Thus, the whole neuronal assembly generating the 20-Hz rebound, rather than a spatially distinct subset, seems to be affected by movement imagination and execution. Figure 5 illustrates the effect of manipulatory finger movements on the 20-Hz activity during ischemiainduced deafferentation and motor paralysis in Subject 1. Immediately after the beginning of ischemia, the subject reported tingling paresthesias which soon became painful. He then developed progressive hypesthesia resulting in a total loss of cutaneous and deep sensation of hand and forearm within 10 min after onset of ischemia. He was still able to move the fingers for another 5 min without having any sensory feedback,
MOTOR IMAGERY AND SENSORIMOTOR CORTEX
205
FIG. 4. (A) Source density plot along the anterior–posterior axis (indicated by the shaded line in B) through the left sensorimotor cortex. N, number of sources; PCS, precentral sulcus; CS, central sulcus. (B) Left-hemisphere sources for the 20-Hz activity of Subject 1 during Rest, superimposed on his magnetic resonance images. The partly overlapping filled white circles represent equivalent current dipoles identified from the bandpassed signals (see Materials and Methods). The black dot indicates the source of the 30-ms evoked response to MN stimuli.
206
SCHNITZLER ET AL.
FIG. 5. Evolution of somatomotor 20-Hz activity during tourniquet-induced ischemia in Subject 1. The signals are from one representative channel above the left sensorimotor hand area. Left half: the thin solid lines indicate the Rest condition prior to and during hypesthesia and the thick dotted lines the Manipulation task prior to ischemia and the attempted Manipulation during ischemia-induced paralysis. Right half: solid lines indicate Rest condition (top) and Manipulation condition (bottom) during ischemia and the dotted lines Rest and Manipulation condition prior to ischemia.
whereafter motor paralysis of small hand and forearm muscles occurred rapidly. Rest and Manipulation runs were performed with the right hand before pressurizing the tourniquet and 5 and 12 min (no sensation) after beginning of ischemia; Manipulation was also performed 21 min after ischemia onset (motor paralysis). During Rest, the characteristic 20-Hz rebound in the contralateral sensorimotor cortex was similar in both preischemic and ischemic recordings. Moreover, the manipulatory finger movements of the right hand abolished the rebound in a very similar manner in both conditions, although the subject was unable to perform any detectable finger movements during ischemia. DISCUSSION The physiological basis of imagined movements has been subject to several functional neuroimaging studies monitoring changes in cerebral blood flow or oxygenation (Roland et al., 1980; Rao et al., 1993; Decety et al., 1994; Lang et al., 1994; Sanes, 1994; Stephan et al., 1995; Grafton et al., 1996). Depending on the technique and the paradigm used, activation has been observed in various cortical and subcortical structures involved in motor behavior. However, the primary motor cortex (MI) was typically silent and only the most recent fMRI studies have suggested that MI is also activated during motor imagery (Leonardo et al., 1995; Roth et al., 1996; Porro et al., 1996; Ersland et al., 1996). In the present study, imagination of self-performed continuous manipulatory finger movements suppressed
the poststimulus somatomotor 20-Hz rhythm. Previous neurophysiological studies of the spatial characteristics and temporal behavior of cortical electromagnetic activity have shown that the 20-Hz rhythm mainly originates in MI (Jasper and Penfield, 1949; Salmelin et al., 1994; Salmelin and Hari, 1995; Conway et al., 1995; Salenius et al., 1997b) whereas the early SEF components are generated in the SI cortex (Hari and Kaukoranta, 1985; Wood et al., 1985; Allison et al., 1991; Baumgartner et al., 1993). Our recordings, which directly reflect neuronal activity, thus allow selective monitoring of the neuronal activity in MI and SI and indicate that MI is involved in motor imagery, in agreement with the recent fMRI reports. In addition to video monitoring of the fingers, we recorded EMG activity of finger extensors during the whole measurement. Although no overt movements were observed, some marginal EMG activity was recorded during motor imagery. However, we can exclude the possibility that this residual motor activity during imagery is responsible for the observed suppression of the 20-Hz activity since the EMG level was lower during Imagination than during Stretch even though the 20-Hz activity was more strongly suppressed during Imagination. Weak EMG activity has frequently been reported during mental motor imagery and is believed to reflect some activation of motor output resulting from increased excitability of spinal motoneurons (Jeannerod, 1995). In fact, Bonnet et al. (1997) demonstrated that the increase in spinal reflex pathway excitability during mental simulation of a movement is almost as strong as during an actual perfor-
MOTOR IMAGERY AND SENSORIMOTOR CORTEX
mance of the same movement. They put forward the interesting hypothesis that during motor imagery an excitatory motor command would be paralleled by an inhibitory motor output from motor cortical areas which actively inhibits the motor effect of the primary motor cortex. An incomplete inhibition would explain the residual EMG activation observed during imagery. Our findings confirm earlier EEG observations that mere thinking about a movement can block the rolandic mu rhythm (Gastaut, 1952; Chatrian et al., 1959). More recent neurophysiological studies have provided some evidence for an involvement of primary motor cortex in motor imagery although alternative explanations could not be ruled out: Beisteiner et al. (1995) recorded similar slow scalp potentials during an imagined and an executed sequence of hand movements. However, the spatial resolution of their scalp EEG recordings was not sufficient to differentiate among activation of MI, premotor areas, and somatosensory cortex. Gandevia and Rothwell (1987) and Hallett et al. (1994), using transcranial motor cortex stimulation, reported increased motor responses during mental simulation of movements but this effect could also be explained by excitability changes at the spinal level. The exact functional role and relevance of somatomotor 20-Hz activity is still unknown although recent studies give evidence that at least part of it is directly related to motor output (Conway et al., 1995; Salenius et al., 1997b). The 20-Hz activity may also reflect functioning of intrinsic cortico-cortical or corticosubcortical neural circuits since it was also affected by movement imagination without motor output. It remains to be shown whether the 20-Hz suppression in the primary motor cortex during motor imagery reflects the representation of a motor program, the inhibitory actions to block movement execution, or the interaction between these processes. The results of the tourniquet-induced ischemia experiment demonstrate that no afferent input from the limbs is required to dampen the 20-Hz rebound in the sensorimotor cortex. Instead, generation of central motor commands alone, without any peripheral motor or reafferent activity, exerts effects on the 20-Hz activity that are similar to those produced by movement execution. Both in the Manipulation task during ischemiainduced motor paralysis and in the Imagination task in the normal condition, motor action is represented but the execution is blocked, in the former case at the peripheral level and in the latter case at a more central level, possibly in the primary motor cortex. Similarly, Ersland et al. (1996) recently observed that a man who had his right arm amputated, and thus had no afferent input from the limb, showed activation in the left motor cortex during imagined movements of the amputated right-hand fingers. Subjects reported that the ability to imagine move-
207
ments fluctuated over time. In line with this, the suppression of the poststimulus 20-Hz bursts during imagery also varied over time within a subject. However, since we did not systematically record the subjects’ rating of their performance, we do not know whether trial-to-trial variability of dampening of 20-Hz rebound is correlated with the subjective fluctuation of performing the imagery task. The common aspect of Manipulation, Small motion, and Imagination conditions seems to be the central representation of motor programs. Only the Manipulation condition is associated with significant cutaneous and muscle afferent activity, leading to a marked modulation of neuronal responses in SI. The lack of activity changes in SI during Imagination is at variance with the findings of a recent fMRI study (Porro et al., 1996), which reported an increased activity level in the postcentral gyrus during motor imagery. The neuronal population generating the 20-Hz somatomotor activity appears to significantly overlap with neurons involved in motor imagery. Thus, our data suggest that intrinsic neuronal circuits of the primary motor cortex are involved in the internal representation of movements irrespective of whether the motor behavior is actually executed or just imagined. ACKNOWLEDGMENTS This work was supported by the Academy of Finland and by the EU’s Human Capital and Mobility programme through the Large Scale Facility BIRCH at the Low Temperature Laboratory, Helsinki University of Technology.
REFERENCES Abbruzzese, G., Trompetto, C., and Schieppati, M. 1996. The excitability of the human motor cortex increases during execution and mental imagination of sequential but not repetitive finger movements. Exp. Brain Res. 111:465–472. Ahonen, A., Ha¨ma¨la¨inen, M. S., Kajola, M. S., Knuutila, J. E. T., Laine, P. P., and Lounasmaa, O. V. 1993. 122-channel SQUID instrument for investigating the magnetic signals from the human brain. Physica Scripta T49:198–205. Allison, T., McCarthy, G., and Wood, C. C. 1991. Potentials evoked in human and monkey cerebral cortex by stimulation of the median nerve: A review of scalp and intracranial recordings. Brain 114: 2465–2503. Baumgartner, C., Doppelbauer, A., Sutherling, W. W., Lindinger, G., and Levesque, M. F. 1993. Somatotopy of human hand somatosensory cortex as studied in scalp EEG. Electroenceph. Clin. Neurophysiol. 88:271–279. Beisteiner, R., Ho¨llinger, P., Lindinger, G., Lang, W., and Berthoz, A. 1995. Mental representations of movements. Brain potentials associated with imagination of hand movements. Electroenceph. Clin. Neurophysiol. 96:183–193. Bonnet, M., Decety, J., Jeannerod, M., and Requin, J. 1997. Mental simulation of an action modulates the excitability of spinal reflex pathways in man. Cogn. Brain Res. 5:221–228. Chatrian, G. E., Petersen, M. C., and Lazarte, J. A. 1959. The blocking of the rolandic wicket rhythm and some central changes
208
SCHNITZLER ET AL.
related to movement. Electroenceph. Clin. Neurophysiol. 11:497– 510. Conway, B. A., Halliday, D. M., Farmer, S. F., Shahani, U., Maas, P., Weir, A-I., and Rosenberg, J. R. 1995. Synchronization between motor cortex and spinal motoneuronal pool during the performance of a maintained motor task in man. J. Physiol. 489:917–924. Crammond, D. J. 1997. Motor imagery: Never in your wildest dream. Trends Neurosci. 20:54–57. Decety, J., Perani, D., Jeannerod, M., Bettinardi, V., Tadary, B., Woods, R., Mazziotta, J. C., and Fazio, F. 1994. Mapping motor representations with positron emission tomography. Nature 371: 600–602. Ersland, L., Rosen, G., Lundervold, A., Smievoll, A. I., Tillung, T., Sundberg, H., and Hugdahl, K. 1996. Phantom limb imagery fingertapping causes primary motor cortex activation: An fMRI study. NeuroReport 7:35–38. Feltz, D. L., and Landers, D. M. 1983. The effects of mental practice on motor skill learning and performance. A meta-analysis. J. Sport Psychol. 5:25–57. Gandevia, S. C., and Rothwell, J. 1987. Knowledge of motor commands and the recruitment of human motoneurons. Brain 110: 1117–1130. Gastaut, H. 1952. Etude electrocorticographique de la re´activite´ des rhythmes rolandiques. Rev. Neurol. 87:176–182. Hallett, M., Fieldman, J., Cohen, L. G., Sadato, N., and PascualLeone, A. 1994. Involvement of primary motor cortex in motor imagery and mental practice. Behav. Brain Sci. 17:210. Ha¨ma¨la¨inen, M., Hari, R., Ilmoniemi, R. J., Knuutila, J., and Lounasmaa, O. V. 1993. Magnetoencephalography—Theory, instrumentation, and applications to noninvasive studies of the working brain. Rev. Mod. Phys. 65:413–497. Hari, R., and Kaukoranta, E. 1985. Neuromagnetic studies of somatosensory system: Principles and examples. Prog. Neurobiol. 24:233– 256. Jasper, H., and Penfield, W. 1949. Electrocorticograms in man: Effect of voluntary movement upon the electrical activity of the precentral gyrus. Arch. Psychiatr. Nervenkr. 183:163–174. Jeannerod, M. 1994. The representing brain: Neural correlates of motor intention and imagery. Behav. Brain Sci. 17:187–245. Jeannerod, M. 1995. Mental imagery in the motor context. Neuropsychology 33:1419–1432. Lang, W., Petit, L., Hollinger, P., Pietrzyk, U., Tzourio, N., Mazoyer, B., and Berthoz, A. 1994. A positron emission tomography study of oculomotor imagery. NeuroReport 5:921–924. Leonardo, M., Fieldman, J., Sadato, N., Campbell, G., Ibanez, V., Cohen, L., Deiber, M. P., Jezzard, P., Pons, T., Turner, R., Le Bihan, D., and Hallett, M. 1995. A functional magnetic-resonance-imaging study of cortical regions associated with motor task execution and motor ideation in humans. Hum. Brain Mapping 3:83–92. Parsons, L. M., Fox, P. T., Downs, J. H., Glass, T., Hirsch, T. B., Martin, C. C., Glass, T., Hirsch, T. B., Martin, C. C., Jerabek, P. A.,
and Lancaster, J. L. 1995. Use of implicit motor imagery for visual shape discrimination as revealed by PET. Nature 375:54–58. Porro, C. A., Francescato, M. P., Cettolo, V., Diamond, M. E., Baraldi, P., Zuiani, C., Bazzochi, M., and di Prampero, P. 1996. Primary motor and sensory cortex activation during motor performance and motor imagery: A functional magnetic resonance imaging study. J. Neurosci. 16:7688–7698. Rao, S. M., Binder, J. R. B., Bandettini, P. A., Hammeke, T. A., Yetkin, F. Z., Jesmanovicz, A., Lisk, L. M., Morris, G. L., Mueller, W. M., and Estkowski, L. D. 1993. Functional magnetic resonance imaging of complex human movements. Neurology 43:2311–2318. Roland, P. E., Larsen, B., Lassen, N. A., and Skinhœj, E. 1980. Supplementary motor area and other areas in organization of voluntary movements in man. J. Neurophysiol. 43:118–136. Roth, R., Decety, J., Raybaudi, M., Massarelli, R., Delon-Martin, C., Segebarth, C., Morand, S., Gemignani, A., and Jeannerod, M. 1996. Possible involvement of primary motor cortex in mentally simulated movement: A functional magnetic resonance imaging study. NeuroReport 7:1280–1284. Salenius, S., Schnitzler, A., Salmelin, R., Jousma¨ki, V., and Hari, R. 1997a. Modulation of human cortical rolandic rhythms during natural sensorimotor tasks. NeuroImage 5:221–228. Salenius, S., Portin, K., Kajola, M., Salmelin, R., and Hari, R. 1997b. Cortical control of human motoneuron firing during isometric contraction. J. Neurophysiol., in press. Salmelin, R., Ha¨ma¨la¨inen, M., Kajola, M., and Hari, R. 1995. Functional segregation of movement-related rhythmic activity in the human brain. NeuroImage 2:237–243. Salmelin, R., and Hari, R. 1994. Spatiotemporal characteristics of sensorimotor MEG rhythms related to thumb movement. Neuroscience 60:537–550. Sanes, J. N. 1994. Neurophysiology of preparation, movement and imagery. Behav. Brain Sci. 17:221–223. Schnitzler, A., Salenius, S., Salmelin, R., Jousma¨ki, V., and Hari, R. 1995a. Involvement of primary sensorimotor cortex in motor imagery: A neuromagnetic study. Soc. Neurosc. Abstr. 21:518. Schnitzler, A., Salmelin, R., Salenius, S., Jousma¨ki, V., and Hari, R. 1995b. Tactile information from the human hand reaches ipsilateral primary somatosensory cortex. Neurosci. Lett. 200:25–28. Sirigu, A., Duhamel, J. R., Cohen, L., Pillon, B., Dubois, B., and Agid, Y. 1996. The mental representation of hand movements after parietal cortex damage. Science 273:1564–1568. Stephan, K. M., Fink, G. R., Passingham, R. E., Silbersweig, D., Silbersweig, D., Ceballos-Baumann, A. O., Frith, C. D., and Frackowiak, R. S. 1995. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J. Neurophysiol. 73:373–386. Wood, C. C., Cohen, D., Cuffin, B. N., Yarita, M., and Allison, T. 1985. Electrical sources in human somatosensory cortex: Identification by combined magnetic and potential recordings. Science 227:1051– 1053.
NEUROIMAGE ARTICLE NO.
Involvement of Primary Motor Cortex in Motor Imagery: A Neuromagnetic Study Alfons Schnitzler,1 Stephan Salenius, Riitta Salmelin, Veikko Jousma¨ki, and Riitta Hari Brain Research Unit, Low Temperature Laboratory, Helsinki University of Technology, P.O. Box 2200, FIN-02015 HUT, Helsinki, Finland Received February 12, 1997
Functional brain imaging studies have indicated that several cortical and subcortical areas active during actual motor performance are also active during imagination or mental rehearsal of movements. Recent evidence shows that the primary motor cortex may also be involved in motor imagery. Using wholescalp magnetoencephalography, we monitored spontaneous and evoked activity of the somatomotor cortex after right median nerve stimuli in seven healthy right-handed subjects while they kinesthetically imagined or actually executed continuous finger movements. Manipulatory finger movements abolished the poststimulus 20-Hz activity of the motor cortex and markedly affected the somatosensory evoked response. Imagination of manipulatory finger movements attenuated the 20-Hz activity by 27% with respect to the rest level but had no effect on the somatosensory response. Slight constant stretching of the fingers suppressed the 20-Hz activity less than motor imagery. The smallest possible, kinesthetically just perceivable finger movements resulted in slightly stronger attenuation of 20-Hz activity than motor imagery did. The effects were observed in both hemispheres but predominantly contralateral to the performing hand. The attempt to execute manipulatory finger movements under experimentally induced ischemia causing paralysis of the hand also strongly suppressed 20-Hz activity but did not affect the somatosensory evoked response. The results indicate that the primary motor cortex is involved in motor imagery. Both imaginative and executive motor tasks appear to utilize the cortical circuitry generating the somatomotor 20-Hz signal. r 1997 Academic Press
INTRODUCTION Motor imagery can be defined as conscious mental rehearsal of a motor act without performing any overt
1 Permanent address: Department of Neurology, Heinrich-HeineUniversity, Moorenstrasse 5, 40225 Du¨sseldorf, Germany.
movement and implies that the subject feels himself executing a given action. Motor imagery thus involves mostly a kinesthetic representation of the action (Jeannerod, 1994). Motor imagery improves motor performance and skill acquisition. Consequently, mental rehearsal of movements is a common technique used by athletes to improve their performance. The beneficial influence of motor imagery on motor performance has previously been reassessed and confirmed in several experiments (Feltz and Landers, 1983), although the neurophysiological mechanisms are not well understood. Interestingly, imagined movements seem to obey the same temporal characteristics and physiological and pathophysiological constraints as executed movements (Sirigu et al., 1996), implying that motor imagery and execution use the same neural representations (for a recent review, see Crammond et al., 1997). Regional cerebral blood flow (rCBF) has been found to increase mainly in the supplementary motor area during imagination of sequential finger movements (Roland et al., 1980). More recent positron emission tomography (PET) (Decety et al., 1994; Stephan et al., 1995; Parsons et al., 1995; Grafton et al., 1996) and functional magnetic resonance imaging (fMRI) studies (Rao et al., 1993) revealed activation of a number of cortical and subcortical areas, including the lateral and medial premotor cortex, anterior cingulate areas, ventral opercular premotor areas, superior and inferior parietal cortex, and posterior cerebellar cortex, but not the primary sensorimotor cortex. In contrast to these findings, the most recent fMRI studies have reported involvement of primary motor cortex during imagined movements in normal subjects (Leonardo et al., 1995; Roth et al., 1996; Porro et al., 1996) and also in a subject with a phantom limb (Ersland et al., 1996). Transcranial magnetic stimulation studies of the motor cortex have pointed in the same direction in showing an increase of motor responses during mental simulation of movements (Gandevia and Rothwell, 1987; Hallett et al., 1994; Abbruzzese et al., 1996); however, such effects could also be explained by excitability changes at the spinal level.
201
1053-8119/97 $25.00 Copyright r 1997 by Academic Press All rights of reproduction in any form reserved.
202
SCHNITZLER ET AL.
In the present study, we assessed cortical activity during imagination and execution of manipulatory finger movements with noninvasive whole-scalp magnetoencephalography. The aim of the study was to find out whether or not the primary somatosensory and motor areas are involved in movement imagination. Experiments were designed to simultaneously monitor evoked activity from the primary somatosensory hand projection cortex (SI) and 20-Hz rhythmic activity from the primary motor cortex (Jasper and Penfield, 1949; Salmelin and Hari, 1994; Salmelin et al., 1995; Conway et al., 1995; Salenius et al., 1997b). We observed a differentially graded suppression of the poststimulus 20-Hz rhythmic activity during execution and imagination of manipulating finger movements, while evoked activity in SI was markedly modulated only by movement execution. Our finding provides evidence that motor imagery is accompanied by changes of neuronal activity within the primary motor cortex and that the same neuronal populations are—to a quantitatively different degree—involved in movement imagination and execution. Results have been presented in abstract form (Schnitzler et al., 1995a).
Imagination, subjects were reminded to imagine and not to perform the task. In one subject the effects of motor imagery on evoked and rhythmic cortical activity were compared with experimentally induced deafferentation and paralysis of the right forearm and hand. This condition bears some analogy to motor imagery: action is represented, but not executed, even though the level at which motor output is blocked is different in the two conditions. Also, since tonic afferent muscular spindle discharges due to increased gamma motoneuron activity during motor imagery could account for improvement of performance (Jeannerod, 1995), deafferentation provides an important reference condition to determine the role of spindle afferent activity during motor imagery. Deafferentation was produced by tourniquet ischemia, increasing the pressure of a blood pressure cuff around the right proximal forearm to 220 mm Hg. The right MN was stimulated at the upper arm by 0.3-ms constant current pulses once every second, with the current strength adjusted before ischemia to produce a slight twitch of the wrist flexors. Neuromagnetic Recording
MATERIALS AND METHODS Subjects and Tasks Seven healthy right-handed members of the laboratory personnel (ages 22–34 years; four females, three males) gave their informed consent to participate in the study. The subject was seated comfortably, with eyes closed, while the right median nerve (MN) was stimulated at the wrist with 0.3-ms constant current pulses once every second. The current strength (4 to 7 mA) was adjusted to produce a slight twitch of the thumb. During stimulation, subjects were instructed to perform eight motor tasks in subsequent runs, each lasting about 5 min. During the Rest condition, the subject sat relaxed without any additional task. In the Manipulation condition, subjects performed continuous finger movements manipulating a small cube between thumb and fingers of one hand. In the Imagination condition, subjects were asked to kinesthetically imagine themselves performing the same manipulating movements without actually executing them. In the Stretch condition, subjects were instructed to keep all finger joints slightly stretched to just overcome the passive flexion force. In the Small motion condition subjects had to perform the smallest possible, just subjectively perceivable, extension and flexion movements of the fingers. All tasks were performed separately with each hand. During all tasks, finger movements were monitored visually using a video camera and electrophysiologically with continuous surface electromyography (EMG) of the finger extensor muscles. If any muscle activity different from Rest was noted on the EMG during
Magnetic brain signals were recorded noninvasively with a Neuromag-122 whole-scalp neuromagnetometer in a magnetically shielded room, while the subject rested his or her head against the helmet-shaped bottom surface of the device (Ahonen et al., 1993). Neuromag-122 contains 122 superconducting planar gradiometers which detect the largest signal above an active cortical area. The signals were recorded with a passband of 0.03–330 Hz, digitized at 1 kHz, and stored for off-line analysis. To determine the position of the head with respect to the sensor array, small currents were led to three indicator coils attached to the scalp to produce magnetic signals. The coil positions were related to the head coordinates, defined by the preauricular points and the nasion, with a 3-D digitizer. Data Analysis Two different analysis approaches were employed to the same data set to simultaneously yield information about the task-dependent modulation of cortical evoked activity and spontaneous rhythms. 1. Somatosensory evoked fields (SEFs). Magnetic signals were digitally low-pass filtered at 140 Hz and about 150 single responses were averaged, time locked to the MN stimuli. A time-varying single-dipole model (Ha¨ma¨la¨inen et al., 1993) was employed to explain the measured signals during the first 60 ms after the stimuli. Superposition of the identified dipoles on individual MRI scans confirmed that these sources reflected activity within the contralateral primary somatosensory cortex, SI (Hari and Kaukoranta, 1985; Wood et
MOTOR IMAGERY AND SENSORIMOTOR CORTEX
al., 1985; Allison et al., 1991; Baumgartner et al., 1993; Schnitzler et al., 1995b). The strongest activity of the SI cortex was compared between tasks. 2. Sensorimotor rhythms. Spectral analysis showed prominent rhythmic activity around 10 and 20 Hz. The temporal spectral evolution method (TSE; Salmelin and Hari, 1994) was employed to calculate the average amplitude level of 20-Hz activity as a function of time, with respect to MN stimuli: the signals were filtered through passbands around 20 Hz, as suggested by spectral analysis. The absolute values were then averaged over about 60 trials and smoothed with a 15-Hz low-pass filter. Somatomotor 20-Hz rhythmic activity shows a characteristic bilateral enhancement (‘‘rebound’’) around 400–700 ms after MN stimuli. This rebound is markedly diminished or even abolished during finger movements (Salenius et al., 1997a). To quantify the modulation of the 20-Hz signals in different tasks, we calculated the vector sum of the TSE signals from the most reactive sensor pair above the sensorimotor hand area (measuring the orthogonal derivatives of the magnetic field at that location). Peak TSE amplitudes in the various tasks were compared with two-tailed t tests for paired differences. Sources of the 20-Hz activity were identified from bandpass-filtered data in 50 poststimulus epochs, 300– 1000 ms after the MN stimulus. Dipolar field patterns were sought at 5-ms intervals, and the strengths, orientations, and three-dimensional locations of the corresponding equivalent current dipoles were determined by means of a least-squares fit to the signals measured by a subset of sensors centered over the hand area. Sources were accepted when they accounted for at least 90% of the measured field variance. The sources can be localized with an accuracy of about 5 mm. RESULTS Figure 1 shows representative 10-s traces of the 20-Hz activity (filtered through 18–25 Hz) above the left sensorimotor cortex of Subject 1 during rest (top) and during the four tasks performed with the right hand. In all conditions the right MN was stimulated electrically once every second. In the rest condition, the stimuli were typically followed by a transient burst-like increase of the 20-Hz rhythmic activity, as described previously in detail (Salenius et al., 1997a). During movement imagination the 20-Hz bursts were dampened, and during actual performance of finger movements (Manipulation and Small motion) they were almost completely suppressed. The reduction of the 20-Hz activity during Imagination varied from intervals of complete suppression to periods of rather small or no effect. Figure 2 illustrates the effects of all tasks on the poststimulus 20-Hz burst, quantified by the TSE
203
FIG. 1. 10-s intervals of spontaneous neuromagnetic activity of Subject 1, recorded above the left sensorimotor hand area during the different tasks (Rest, Imagination, Manipulation, Stretch, Small motion). The signals were bandpass filtered through 18–25 Hz. Dashed vertical lines indicate right median nerve stimuli delivered once per second. Note the bursts following each MN stimulus and the suppression of these bursts during the tasks.
method, and on the SEFs to right MN stimulation, recorded from the left sensorimotor cortex. In the rest condition, the rebound of the 20-Hz activity started about 200 ms after stimulus onset and peaked at around 450 ms. Overt manipulatory and smallest possible finger movements abolished the 20-Hz rebound in this subject almost completely. Movement imagination led to a 50% reduction, and simple static finger stretch led to only a slight decrease of the rebound. These effects were most prominent when the task was performed with the stimulated (right) hand. SEFs of the contralateral SI cortex were markedly affected by manipulatory movements (Manipulation) which caused activation of cutaneous and deep afferents. No discernible changes were observed in SI responses during the other conditions (Imagination, Stretch, and Small Motion). Electromyographic activity of finger extensors was largest during Manipulation and only marginal
204
SCHNITZLER ET AL.
FIG. 2. Task-related effects on somatomotor 20-Hz activity, quantified with the TSE method, and EMG and SEFs recorded above the left sensorimotor hand cortex in Subject 1, with the EMG activity plotted below. The different tasks (Rest, Imagination, Manipulation, Stretch, Small motion) are illustrated by different line styles. The left and right columns show the results for the tasks performed with the right and left hands, respectively. (Top row) Temporal evolution of 20-Hz activity. Sixty epochs (each 1 s in duration) were averaged with respect to right MN stimuli. (Middle row) Rectified and averaged EMG of right finger extensors recorded simultaneously with the cortical signals. (Bottom row) SEFs to stimulation of the right MN.
during Imagination. However, there was no direct relation between reduction of the 20-Hz rebound and muscular activity: although EMG activity was higher for Stretch than for Imagination, the 20-Hz activity was more strongly suppressed during Imagination. Figure 3 illustrates the mean (1SEM) effects of the different tasks on the 20-Hz rebound compared with the rest condition. The poststimulus 20-Hz rebound was reduced statistically significantly during all tasks (P , 0.005). The most pronounced suppression was observed during actual movements of the right hand (Manipulation 45% and Small motion 39%), with the
FIG. 3. Mean (1 SEM), seven subjects) amplitudes of the 20-Hz rebound for the different tasks relative to the rest condition (5100%).
smallest effect occurring during simple static muscular activity (Stretch 18%) and an intermediate effect (27%) during Imagination. The 20-Hz activity was also dampened, but to a lesser degree (11–37%; P , 0.01), when the tasks were performed with the left hand. Figure 4 illustrates the current dipole cluster determined from the 20-Hz bursts during Rest in Subject 1, superimposed on the surface image of his brain. During all tasks, the sources covered the same area in the hand motor cortex. Source density plot along the anterior– posterior axis shows that the 20-Hz activity was generated in the motor cortex. Similar results were obtained in the other subjects. Thus, the whole neuronal assembly generating the 20-Hz rebound, rather than a spatially distinct subset, seems to be affected by movement imagination and execution. Figure 5 illustrates the effect of manipulatory finger movements on the 20-Hz activity during ischemiainduced deafferentation and motor paralysis in Subject 1. Immediately after the beginning of ischemia, the subject reported tingling paresthesias which soon became painful. He then developed progressive hypesthesia resulting in a total loss of cutaneous and deep sensation of hand and forearm within 10 min after onset of ischemia. He was still able to move the fingers for another 5 min without having any sensory feedback,
MOTOR IMAGERY AND SENSORIMOTOR CORTEX
205
FIG. 4. (A) Source density plot along the anterior–posterior axis (indicated by the shaded line in B) through the left sensorimotor cortex. N, number of sources; PCS, precentral sulcus; CS, central sulcus. (B) Left-hemisphere sources for the 20-Hz activity of Subject 1 during Rest, superimposed on his magnetic resonance images. The partly overlapping filled white circles represent equivalent current dipoles identified from the bandpassed signals (see Materials and Methods). The black dot indicates the source of the 30-ms evoked response to MN stimuli.
206
SCHNITZLER ET AL.
FIG. 5. Evolution of somatomotor 20-Hz activity during tourniquet-induced ischemia in Subject 1. The signals are from one representative channel above the left sensorimotor hand area. Left half: the thin solid lines indicate the Rest condition prior to and during hypesthesia and the thick dotted lines the Manipulation task prior to ischemia and the attempted Manipulation during ischemia-induced paralysis. Right half: solid lines indicate Rest condition (top) and Manipulation condition (bottom) during ischemia and the dotted lines Rest and Manipulation condition prior to ischemia.
whereafter motor paralysis of small hand and forearm muscles occurred rapidly. Rest and Manipulation runs were performed with the right hand before pressurizing the tourniquet and 5 and 12 min (no sensation) after beginning of ischemia; Manipulation was also performed 21 min after ischemia onset (motor paralysis). During Rest, the characteristic 20-Hz rebound in the contralateral sensorimotor cortex was similar in both preischemic and ischemic recordings. Moreover, the manipulatory finger movements of the right hand abolished the rebound in a very similar manner in both conditions, although the subject was unable to perform any detectable finger movements during ischemia. DISCUSSION The physiological basis of imagined movements has been subject to several functional neuroimaging studies monitoring changes in cerebral blood flow or oxygenation (Roland et al., 1980; Rao et al., 1993; Decety et al., 1994; Lang et al., 1994; Sanes, 1994; Stephan et al., 1995; Grafton et al., 1996). Depending on the technique and the paradigm used, activation has been observed in various cortical and subcortical structures involved in motor behavior. However, the primary motor cortex (MI) was typically silent and only the most recent fMRI studies have suggested that MI is also activated during motor imagery (Leonardo et al., 1995; Roth et al., 1996; Porro et al., 1996; Ersland et al., 1996). In the present study, imagination of self-performed continuous manipulatory finger movements suppressed
the poststimulus somatomotor 20-Hz rhythm. Previous neurophysiological studies of the spatial characteristics and temporal behavior of cortical electromagnetic activity have shown that the 20-Hz rhythm mainly originates in MI (Jasper and Penfield, 1949; Salmelin et al., 1994; Salmelin and Hari, 1995; Conway et al., 1995; Salenius et al., 1997b) whereas the early SEF components are generated in the SI cortex (Hari and Kaukoranta, 1985; Wood et al., 1985; Allison et al., 1991; Baumgartner et al., 1993). Our recordings, which directly reflect neuronal activity, thus allow selective monitoring of the neuronal activity in MI and SI and indicate that MI is involved in motor imagery, in agreement with the recent fMRI reports. In addition to video monitoring of the fingers, we recorded EMG activity of finger extensors during the whole measurement. Although no overt movements were observed, some marginal EMG activity was recorded during motor imagery. However, we can exclude the possibility that this residual motor activity during imagery is responsible for the observed suppression of the 20-Hz activity since the EMG level was lower during Imagination than during Stretch even though the 20-Hz activity was more strongly suppressed during Imagination. Weak EMG activity has frequently been reported during mental motor imagery and is believed to reflect some activation of motor output resulting from increased excitability of spinal motoneurons (Jeannerod, 1995). In fact, Bonnet et al. (1997) demonstrated that the increase in spinal reflex pathway excitability during mental simulation of a movement is almost as strong as during an actual perfor-
MOTOR IMAGERY AND SENSORIMOTOR CORTEX
mance of the same movement. They put forward the interesting hypothesis that during motor imagery an excitatory motor command would be paralleled by an inhibitory motor output from motor cortical areas which actively inhibits the motor effect of the primary motor cortex. An incomplete inhibition would explain the residual EMG activation observed during imagery. Our findings confirm earlier EEG observations that mere thinking about a movement can block the rolandic mu rhythm (Gastaut, 1952; Chatrian et al., 1959). More recent neurophysiological studies have provided some evidence for an involvement of primary motor cortex in motor imagery although alternative explanations could not be ruled out: Beisteiner et al. (1995) recorded similar slow scalp potentials during an imagined and an executed sequence of hand movements. However, the spatial resolution of their scalp EEG recordings was not sufficient to differentiate among activation of MI, premotor areas, and somatosensory cortex. Gandevia and Rothwell (1987) and Hallett et al. (1994), using transcranial motor cortex stimulation, reported increased motor responses during mental simulation of movements but this effect could also be explained by excitability changes at the spinal level. The exact functional role and relevance of somatomotor 20-Hz activity is still unknown although recent studies give evidence that at least part of it is directly related to motor output (Conway et al., 1995; Salenius et al., 1997b). The 20-Hz activity may also reflect functioning of intrinsic cortico-cortical or corticosubcortical neural circuits since it was also affected by movement imagination without motor output. It remains to be shown whether the 20-Hz suppression in the primary motor cortex during motor imagery reflects the representation of a motor program, the inhibitory actions to block movement execution, or the interaction between these processes. The results of the tourniquet-induced ischemia experiment demonstrate that no afferent input from the limbs is required to dampen the 20-Hz rebound in the sensorimotor cortex. Instead, generation of central motor commands alone, without any peripheral motor or reafferent activity, exerts effects on the 20-Hz activity that are similar to those produced by movement execution. Both in the Manipulation task during ischemiainduced motor paralysis and in the Imagination task in the normal condition, motor action is represented but the execution is blocked, in the former case at the peripheral level and in the latter case at a more central level, possibly in the primary motor cortex. Similarly, Ersland et al. (1996) recently observed that a man who had his right arm amputated, and thus had no afferent input from the limb, showed activation in the left motor cortex during imagined movements of the amputated right-hand fingers. Subjects reported that the ability to imagine move-
207
ments fluctuated over time. In line with this, the suppression of the poststimulus 20-Hz bursts during imagery also varied over time within a subject. However, since we did not systematically record the subjects’ rating of their performance, we do not know whether trial-to-trial variability of dampening of 20-Hz rebound is correlated with the subjective fluctuation of performing the imagery task. The common aspect of Manipulation, Small motion, and Imagination conditions seems to be the central representation of motor programs. Only the Manipulation condition is associated with significant cutaneous and muscle afferent activity, leading to a marked modulation of neuronal responses in SI. The lack of activity changes in SI during Imagination is at variance with the findings of a recent fMRI study (Porro et al., 1996), which reported an increased activity level in the postcentral gyrus during motor imagery. The neuronal population generating the 20-Hz somatomotor activity appears to significantly overlap with neurons involved in motor imagery. Thus, our data suggest that intrinsic neuronal circuits of the primary motor cortex are involved in the internal representation of movements irrespective of whether the motor behavior is actually executed or just imagined. ACKNOWLEDGMENTS This work was supported by the Academy of Finland and by the EU’s Human Capital and Mobility programme through the Large Scale Facility BIRCH at the Low Temperature Laboratory, Helsinki University of Technology.
REFERENCES Abbruzzese, G., Trompetto, C., and Schieppati, M. 1996. The excitability of the human motor cortex increases during execution and mental imagination of sequential but not repetitive finger movements. Exp. Brain Res. 111:465–472. Ahonen, A., Ha¨ma¨la¨inen, M. S., Kajola, M. S., Knuutila, J. E. T., Laine, P. P., and Lounasmaa, O. V. 1993. 122-channel SQUID instrument for investigating the magnetic signals from the human brain. Physica Scripta T49:198–205. Allison, T., McCarthy, G., and Wood, C. C. 1991. Potentials evoked in human and monkey cerebral cortex by stimulation of the median nerve: A review of scalp and intracranial recordings. Brain 114: 2465–2503. Baumgartner, C., Doppelbauer, A., Sutherling, W. W., Lindinger, G., and Levesque, M. F. 1993. Somatotopy of human hand somatosensory cortex as studied in scalp EEG. Electroenceph. Clin. Neurophysiol. 88:271–279. Beisteiner, R., Ho¨llinger, P., Lindinger, G., Lang, W., and Berthoz, A. 1995. Mental representations of movements. Brain potentials associated with imagination of hand movements. Electroenceph. Clin. Neurophysiol. 96:183–193. Bonnet, M., Decety, J., Jeannerod, M., and Requin, J. 1997. Mental simulation of an action modulates the excitability of spinal reflex pathways in man. Cogn. Brain Res. 5:221–228. Chatrian, G. E., Petersen, M. C., and Lazarte, J. A. 1959. The blocking of the rolandic wicket rhythm and some central changes
208
SCHNITZLER ET AL.
related to movement. Electroenceph. Clin. Neurophysiol. 11:497– 510. Conway, B. A., Halliday, D. M., Farmer, S. F., Shahani, U., Maas, P., Weir, A-I., and Rosenberg, J. R. 1995. Synchronization between motor cortex and spinal motoneuronal pool during the performance of a maintained motor task in man. J. Physiol. 489:917–924. Crammond, D. J. 1997. Motor imagery: Never in your wildest dream. Trends Neurosci. 20:54–57. Decety, J., Perani, D., Jeannerod, M., Bettinardi, V., Tadary, B., Woods, R., Mazziotta, J. C., and Fazio, F. 1994. Mapping motor representations with positron emission tomography. Nature 371: 600–602. Ersland, L., Rosen, G., Lundervold, A., Smievoll, A. I., Tillung, T., Sundberg, H., and Hugdahl, K. 1996. Phantom limb imagery fingertapping causes primary motor cortex activation: An fMRI study. NeuroReport 7:35–38. Feltz, D. L., and Landers, D. M. 1983. The effects of mental practice on motor skill learning and performance. A meta-analysis. J. Sport Psychol. 5:25–57. Gandevia, S. C., and Rothwell, J. 1987. Knowledge of motor commands and the recruitment of human motoneurons. Brain 110: 1117–1130. Gastaut, H. 1952. Etude electrocorticographique de la re´activite´ des rhythmes rolandiques. Rev. Neurol. 87:176–182. Hallett, M., Fieldman, J., Cohen, L. G., Sadato, N., and PascualLeone, A. 1994. Involvement of primary motor cortex in motor imagery and mental practice. Behav. Brain Sci. 17:210. Ha¨ma¨la¨inen, M., Hari, R., Ilmoniemi, R. J., Knuutila, J., and Lounasmaa, O. V. 1993. Magnetoencephalography—Theory, instrumentation, and applications to noninvasive studies of the working brain. Rev. Mod. Phys. 65:413–497. Hari, R., and Kaukoranta, E. 1985. Neuromagnetic studies of somatosensory system: Principles and examples. Prog. Neurobiol. 24:233– 256. Jasper, H., and Penfield, W. 1949. Electrocorticograms in man: Effect of voluntary movement upon the electrical activity of the precentral gyrus. Arch. Psychiatr. Nervenkr. 183:163–174. Jeannerod, M. 1994. The representing brain: Neural correlates of motor intention and imagery. Behav. Brain Sci. 17:187–245. Jeannerod, M. 1995. Mental imagery in the motor context. Neuropsychology 33:1419–1432. Lang, W., Petit, L., Hollinger, P., Pietrzyk, U., Tzourio, N., Mazoyer, B., and Berthoz, A. 1994. A positron emission tomography study of oculomotor imagery. NeuroReport 5:921–924. Leonardo, M., Fieldman, J., Sadato, N., Campbell, G., Ibanez, V., Cohen, L., Deiber, M. P., Jezzard, P., Pons, T., Turner, R., Le Bihan, D., and Hallett, M. 1995. A functional magnetic-resonance-imaging study of cortical regions associated with motor task execution and motor ideation in humans. Hum. Brain Mapping 3:83–92. Parsons, L. M., Fox, P. T., Downs, J. H., Glass, T., Hirsch, T. B., Martin, C. C., Glass, T., Hirsch, T. B., Martin, C. C., Jerabek, P. A.,
and Lancaster, J. L. 1995. Use of implicit motor imagery for visual shape discrimination as revealed by PET. Nature 375:54–58. Porro, C. A., Francescato, M. P., Cettolo, V., Diamond, M. E., Baraldi, P., Zuiani, C., Bazzochi, M., and di Prampero, P. 1996. Primary motor and sensory cortex activation during motor performance and motor imagery: A functional magnetic resonance imaging study. J. Neurosci. 16:7688–7698. Rao, S. M., Binder, J. R. B., Bandettini, P. A., Hammeke, T. A., Yetkin, F. Z., Jesmanovicz, A., Lisk, L. M., Morris, G. L., Mueller, W. M., and Estkowski, L. D. 1993. Functional magnetic resonance imaging of complex human movements. Neurology 43:2311–2318. Roland, P. E., Larsen, B., Lassen, N. A., and Skinhœj, E. 1980. Supplementary motor area and other areas in organization of voluntary movements in man. J. Neurophysiol. 43:118–136. Roth, R., Decety, J., Raybaudi, M., Massarelli, R., Delon-Martin, C., Segebarth, C., Morand, S., Gemignani, A., and Jeannerod, M. 1996. Possible involvement of primary motor cortex in mentally simulated movement: A functional magnetic resonance imaging study. NeuroReport 7:1280–1284. Salenius, S., Schnitzler, A., Salmelin, R., Jousma¨ki, V., and Hari, R. 1997a. Modulation of human cortical rolandic rhythms during natural sensorimotor tasks. NeuroImage 5:221–228. Salenius, S., Portin, K., Kajola, M., Salmelin, R., and Hari, R. 1997b. Cortical control of human motoneuron firing during isometric contraction. J. Neurophysiol., in press. Salmelin, R., Ha¨ma¨la¨inen, M., Kajola, M., and Hari, R. 1995. Functional segregation of movement-related rhythmic activity in the human brain. NeuroImage 2:237–243. Salmelin, R., and Hari, R. 1994. Spatiotemporal characteristics of sensorimotor MEG rhythms related to thumb movement. Neuroscience 60:537–550. Sanes, J. N. 1994. Neurophysiology of preparation, movement and imagery. Behav. Brain Sci. 17:221–223. Schnitzler, A., Salenius, S., Salmelin, R., Jousma¨ki, V., and Hari, R. 1995a. Involvement of primary sensorimotor cortex in motor imagery: A neuromagnetic study. Soc. Neurosc. Abstr. 21:518. Schnitzler, A., Salmelin, R., Salenius, S., Jousma¨ki, V., and Hari, R. 1995b. Tactile information from the human hand reaches ipsilateral primary somatosensory cortex. Neurosci. Lett. 200:25–28. Sirigu, A., Duhamel, J. R., Cohen, L., Pillon, B., Dubois, B., and Agid, Y. 1996. The mental representation of hand movements after parietal cortex damage. Science 273:1564–1568. Stephan, K. M., Fink, G. R., Passingham, R. E., Silbersweig, D., Silbersweig, D., Ceballos-Baumann, A. O., Frith, C. D., and Frackowiak, R. S. 1995. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J. Neurophysiol. 73:373–386. Wood, C. C., Cohen, D., Cuffin, B. N., Yarita, M., and Allison, T. 1985. Electrical sources in human somatosensory cortex: Identification by combined magnetic and potential recordings. Science 227:1051– 1053.