Brain potentials associated with movement in traumatic brain injury.

Brain Potentials Associated with Movement in Traumatic Brain Injury Rehabilitation of the patient with neurological dysfunction is often accompanied by attempts to find neurophysiological correlates of motor dysfunction in order for the rehabilitation professional to base training programs on firm scientific evidence. This article discusses the technique of recording movement-related brain potentials (MRPs) and its use in assessing and predicting motor dysfunction in individuals with traumatic brain injury (TBI).

Unfortunately, the number of existing noninvasive physiological measures that are sensitive to motor dysfunctions (ie, measures that can detect abnormalities, in particular, those in brain structures or brain processes) and ca thus help direct rehabilitative efforts is scarce. This scarcity is partially due to the fact that many of the techniques that hold promise for clinical use are still in the early stages of development. Many of these neurophysiologically focused assessment techniques have been used in clinical settings to investigate differences between groups of patients and control subjects. The narrow focus of some of these measures has also hindered their acceptance for clinical practice. The information provided by sensory evoked potentials or by investigation of the spinal reflex, for example, is limited and often difficult to interpret and thus may discourage clinicians from using these measures to plan therapy. Disorders of movement may result from underlying abnormalities in attentional, sensory, or motor processes or, most likely, from a combination of these mechanisms. This article examines whether electrocortical correlates of preparation for and execution of simple movements can indicate whether there is a breakdown in the sensorimotor processes subserving voluntary movement.

Movement-Related Potentials

Event-related brain potentials (ERPs), recorded through averaging techniques, are time-locked to sensory or motor events. Thus, ERPs display the brain activity specific to these events. Movement-related potential studies, which are part of the ERP field, have typically investigated an epoch (time interval) of 1.5 to 2 seconds preceding the electromyographic (EMG) burst in simple unilateral ballistic finger or hand movement in addition to a short period (approximately 500 milliseconds) following muscle activity. The widespread, slow cortical negativity preceding such simple movement has been well documented in healthy adults. [1-5] Although there has not always been agreement regarding the terminology of this cortical activity (eg, terms such as "motor potential," [1] "readiness potential," [2] "movement-related potential," "movement-related cortical potential," [4] and "Bereitschafts-potential" [5] have been used inter-changeably), there is a consensus as to the main components of a waveform. These components, first reported by Kornhuber and Deecke, [6,7] are as follows: (1) a slow negativity preceding the movement by about 1 second (N1), (2) a small positivity just prior to EMG activity (P1), (3) a sharp negativity peaking, about the time of movement initiation (N2), and (4) a final slow and large positivity (P2). Figure 1 displays the MRP associated with right-hand finger flexion of a healthy adult subject and illustrates the main components.

It can now be determined with relative certainty that the MRP is more than just a cortical correlate of muscle activation. The amplitude of the readiness potential has been shown to reflect planning processes closely linked to the complexity of the associated movement. [8,9] Various authors [9,10] have suggested that the supplementary motor area (SMA) in the mesial wall of the frontal lobe (ie, the fold between the two hemispheres) is strongly involved in nonspecific preparatory processes for movement. This suggestion is supported by the fact that the surface-recorded readiness potential is most prominent over the SMA (between cortical electrodes Cz and Fz). Whereas most evoked potential techniques investigate the processing of afferent signals (eg, the sensory system), MRPs provide information regarding the planning and execution of motor acts and sensory processes. More specifically, visual, auditory, and somatosensory evoked potentials can be visible in the MRP traces. Like other electroencephalographic techniques, the MRP paradigm can also indicate the relative involvement of different brain areas in the movement process.

Recent developments may make the MRP technique a useful clinical tool for the assessment of movement control in general and of movement disorders in specific. As research on MRPs has evolved, more complicated paradigms involving different types of movements have been used in place of the simple unilateral finger flexion movements originally investigated. Consequently, various components of brain activity associated with movement have been identified. A comprehensive review of research on MRPs can be found elsewhere. [10] This article attempts to give the reader a basic understanding of the technique and to review those findings that may be most relevant to the investigation of movement dysfunction associated with brain injuries.

Functional Significance of

Movement-Related Potentials

The conventionally held notions of slow negative shift preceding the motor act and of contralateral predominance of the latter part of this potential (ie, larger potentials over the contralateral hemisphere) have been contested by various investigators. Brunia and Vingerhoets [11, 12] investigated brain potentials that preceded unilateral plantar flexion of the right foot. These brain potentials demonstrated larger readiness potential amplitudes on the ipsilateral (right) side of the brain. Kristeva and Deecke [13] showed a significantly larger Bereitschaftspotential (readiness potential) amplitude over the right precentral area prior to bilateral simultaneous finger movements. Increased effort in the right hemisphere controlling the nondominant (left) hand was offered as a possible explanation for the latter findings. The reasons for the uncommon distribution of the readiness potentials in these examples are not apparent. The functional mechanisms underlying the slow negative premovement potentials are not yet fully understood. I believe the amplitude of the slow negative wave preceding movement may not reflect neural processes specific to the motor command, but rather a cognitive process associated with the allocation of attention necessary to complete the motor act.

Karrer and co-workers [14, 15] observed that, unlike the slow negative shift recorded in healthy adults, slow positive waves preceded movements in young children and in mentally retarded adults. These investigators suggested that inhibitory processes associated with a less-advanced level of motor control lead to premovement positivity. In the case of children, they argue, this premovement positivity will gradually shift to the more familiar premotor negativity. This interpretation, however, is in contrast to the findings of Kristeva and Deecke [13] which showed that the "less-skilled" hemisphere displayed increased, rather than reduced, premotor negativity.

It could also be proposed that a higher degree of differentiation, secondary to the development of the neocortex, results in a shift of movement control processes to the cortex. This proposal seems particularly reasonable for single-joint peripheral movements of the type used in most MRP studies. Neuronal activity related to attentional processes used to monitor a motor act, rather than to control the motor coordination of the actual movement, may reflect the extent of cortical differentiation and appear as premotor negativity. Conversely, I believe that reduced cortical involvement in the control of movement or increased subcortical contribution within the control network will translate to premotor positive surface activity.

Promising developments in research on MRPs come from the study of longer duration movements and in particular from the investigation of more complex movements. Deecke et al [10] had their subjects lower a ballpoint pen to a digitized surface and track a visual or tactile stimulus that changed direction every second (thus making the task predictable in the latency domain, but not in the spatial domain). An intriguing finding of this study was the large-amplitude positive deflections occurring at the time of tracking. These positive deflections, sandwiched between the negative potentials associated with the initiation as well as the change in direction of the movement, were termed by the authors "relaxation potentials." Deecke et al [10] contended that surface positivity reflects cortical relaxation that follows the active participation of the cortex in processes such as monitoring and specifying changes in an ongoing movement.

Positive surface activity associated with sustained and skilled movements were also demonstrated to occur during handwriting. [16] In contrast, a sustained negative shift has been observed in a slow aiming task in which the index finger had to reac a specified amplitude of displacement or force. [17] This finding again suggests that attentional requirements for the movement or for active "cerebral monitoring" are necessary for movement-related negativity, whereas positive potentials are often observed with the execution of both simple and very skilled automatic movements. The study of longer-duration movements, therefore, has resulted in the identification of "true" MRPs (ie, potentials associated with the control of the ongoing movement) to accompany the already explored premotor and postmotor (termination) brain potentials. Figure 2 schematically describes the major components associated with ballistic as well as sustained movements. Figure 2, however, does not offer a complete description of reported MRP components (see article by Deecke et al [10] for more information).

Movement-Related Potentials in

a Clinical Population

The recent expansion of the MRP paradigm to sustained and more complex movements has not yet included studies of clinical populations. Even the clinical utilization of this tool in its more basic form (investigation of single-joint, unilateral ballistic movements) has been limited. Yet, MRP studies documenting brain activity associated with voluntary movements in cerebellar [18] and basal ganglia [19, 20] disorders show that the possibility of detecting cerebral abnormalities associated with movement disorders is good. For example, Shibasaki et al [18] found MRPs sensitive to the locus of cerebellar lesions, showing normal readiness potential in cerebellar cortex, but not dentate nucleus, involvement.

In patients with parkinsonian akinesia, [19] an abnormal MRP associated with voluntary finger movement was observed. Movement of the right (nonparetic) finger was preceded by a grossly enhanced contralateral negative brain potential, resulting in a pronounced asymmetry in the readiness potential. The brain activity preceding left (paretic) finger movement, in contrast, showed an uncharacteristic symmetry brought about by similar excitation in the ipsilateral (nonparetic) precentral region and in the contralateral (paretic) precentral region. Deecke et al [10] have noted that people with Parkinson's disease were the only subjects demonstrating positive readiness activity other than young children and mentally retarded adults.

The MRP is a multicomponent complex that appears to include contributions from the sensory, motor, and attentional domains. As such, it can provide valuable information concerning the exact nature of movement impairments following brain trauma (ie, whether they are of sensory, motor, or attentional origin). However, to my knowledge, MRPs have not been investigated in TBI or stroke patients. Co-contraction of antagonist muscle grops, as well as unclear onset of the EMG activity in these patients, may have deterred investigators from pursuing this type of research. A paradigm utilizing a mechanical trigger, rather than an EMG trigger, for the averaging of brain potentials, however, should enable researchers to record MRPs in paretic subjects as well. The study reported here was an initial effort to record MRPs in tBI patients and to compare them, as a group and individually, with healthy controls.

Method

Subjects

Movement-related brain potentials associated with finger flexion and with elbow flexion and extension were studied in five postacute TBI patients (an average of 5.2 years postacute [SD=2.99, range=1.5-8.0]) and in four healthy control subjects. The TBI patients ranged in age between 18 and 26 years (X=22.8, SD=3.4). Three TBI patients were left hemiparetic, and two were right hemiparetic. Despite increased muscular tone and poor motor control on the involved side, all TBI patients were able to perform the experimental tasks. The control group subjects who ranged in age between 29 and 38 years (X=31.5, SD=4.36), used their right hand in the three experimental conditions. Informed consent was obtained from all subjects after the procedures had been fully explained to them.

Procedure

For both practical and theoretical reasons, a goal-directed movement was chosen for the experimental paradigm. Thus, the subjects performed a motor task instead of merely activating a muscle. The subjects were asked to squeeze a mechanical trigger by flexing their index finger (condition 1) or by a slight flexion (condition 2) or extension (condition 3) of the elbow. In conditions 2 and 3, the trigger was secured to a height-adjustable table with the subject's forearm resting in a rotating saddle. The table with the trigger apparatus was placed at shoulder height, with the trigger at the distal end of the subject's arm. Because the movement task in each condition required only a slight movement to produce a deflection of the trigger (3-4 mm), the muscle action was regarded as "quasi-isometric."

Subjects performed two blocks of 60 brisk, discrete, unrejected (artifact rejection function, eliminating noise from movement or from muscle activity, was used in the recording process) movements in each arm and for each condition. Only the right arm was assessed in the control group. Seven channels of EEG and one surface EMG channel were recorded and averaged on-line (Nicolet Pathfinder 1 computer (*1). Silver electrodes were place on the scalp according to the 10/20 international system at positions Cz, FCz (midway betweeen Cz and Fz), C3' and C4' (just anterior to C3 and C4, respectively), and P3 and P4. Cz is a vertex electrode positioned halfway between the inion and the nasion. C3 and C4 mark the hand area in the left and right hemispheres, respectively (roughly 7.6 cm lateral to the vertex). P3 and P4 were attached over the left and right parietal cortices, respectively (Fig. 3). All electrodes were referenced against linked ears (Al+r), and a ground electrode was placed on the forehead (Fpz). Electrode impedance was kept under 3 k[Omega]. Rectified and integrated EMG signals from the flexor digitorum, biceps brachii, and triceps brachii muscles were recorded when finger flexions, elbow flexions, and elbow extensions, respectively, were performed.

Two seconds of movement-related brain activity was analyzed, 1.5 seconds prior to and 0.5 second following the trigger squeeze. During the data-collection, periods, the subjects sat on a chair, fixating on a mark placed at eye level on a wall 1 m in front of them. They were asked to relax, minimize associated movements, and activate the trigger at irregular intervals of 2 to 4 seconds. The order of the three experimental conditions was randomized across subjects, and, in the case of the TBI patients, the nonparetic arm was tested before the paretic arm.

Data Analysis

The data from the four control subjects were average for each experimental condition and plotted against the respective averages of three left-hemiparetic TBI patients (Fig. 3). In addition, the averaged MRP traces from each subject were further analyzed for the amplitudes of three components. One component was the MRP amplitude difference between the onset of the slow negative wave and the EMG onset, thus depicting the flow readiness activity. The second component measured was the difference between the amplitude at the point of trigger activation and the amplitude at the negative peak (N2). The third component measured was the amplitude difference between peak negativity (N2) and the following postrigger positive deflection P2). The Table compares these components in the TBI group (for both paretic and nonparetic hands) with those of the control subjects in the three experimental tasks. Unusual MRP traces for two of the left-hemiparetic TBI patients are demonstrated in Figure 4.

The final analysis of the data was a correlation analysis of four cortical deviations (all referenced to linked ears): C3 versus C4 (frontal coherence), P3 versus P4 (parietal coherence), C3 versus P3 (left-hemisphere coherence), and C4 versus P4 (right-hemisphere coherence). The derived correlations were plotted as profiles of individual subject. Figure 5 illustrates the profiles of cross-cortical corelation computed from MRPs associated with elbow extension in the five TBI patients and the four control subjects.

Results

Amplitude Analysis

Figure 3 shows that very little pre-movement cortical activity occurred in the TBI patients when their paretic arm was tested in experimental conditions 1 (finger flexion) and 2 (elbow flexion). In condition 3 (elbow extension), however, late, yet prominent, cortical activity preceded the trigger activation. The TBI patients' nonparetic arm movement, although exhibiting more pronounced cortical preparatory activity, showed MRPs with different shapes and lower amplitudes than the respective MRPs in the control group. Unlike the parkinsonian subjects in the Deecke et al[19] study, who demonstrated enhanced contralateral negative potentials associated with right (nonparetic) finger movement, the contralateral negativity in the TBI patients in this study did not appear to be enhanced. Their contralateral negativity differed from the respective negativity over the ipsilateral (paretic) motor area only during the 100- to 200-millisecond period immediately preceding the trigger activation (Table). For the midfrontal locations (FCz, Cz), the slow negative shift preceding the movement was more pronounced in the control subjects as compared with the TBI patients in all three experimental conditions. Because of the small group size, however, few of these comparisons were statistically significant. Similarly, the N2-P2 wave was larger in the control subjects than in the TBI patients. There were no obvious differences, however, between the TBI patients and the control subjects with regard to readiness potentials recorded over the motor strip (C3', C4'). In these electrode configurations, the only consistent observation was the more pronounced N2-P2 amplitude in the control subjects.

The amplitude difference between time of trigger activation and peak negativity in the control subjects was relatively small because these two components often coincided. In the parietal lobe, however, cortical negativity peaked later in the control group than in the TBI group. This result was manifested in larger N2-trigger activation amplitude in the parietal leads (mainly P4) in that group. The TBI group had unusually large potentials shifts between N2 and trigger activation for the paretic arm movement in the extension condition. This finding, best demonstrated in Figure 3, is a result of an early peak of cortical negativity at the onset of EMG activity. In the control subjects, the peak negativity or motor potential was synchronized with the trigger or followed it (Fig. 3). In contrast, paretic hand movements in the TBI group were usually characterized by an early negative peak preceding the trigger by a few hundred milliseconds. This peak was followed by a gradual decrease in amplitude that ended at P2 (with the exception of the elbow extension condition described earlier).

Waveform Abnormality

Unlike the control subjects' similar and well-synchronized activity over homologous cortical sites (eg, C3 and C4, P3 and P4), the TBI patients frequently demonstrated unsynchronized MRP traces over homologous sites (Fig. 4). This unusual activity was associated with movement of either the paretic or the nonparetic arm and often occurred in only one experimental condition. For example, subject DI (left hemiparetic) showed uncorrelated left and right parietal activity. In right elbow extension, there was sustained negativity over her left parietal lobe, the peak of which was synchronized with P2. At the same time, there was slow positive deflection over the right (paretic) hemisphere. This phenomenon was reversed for right elbow flexion. The MRP records of another left-hemiparetic TBI patient (subject ST) showed that, for left (paretic) finger flexion, the gradual negative shift over the right frontal lobe (C4') ended long before the trigger activation and turned into a positive slope. This result was not paralleled by the left hemisphere, which demonstrated the more familiar negative polarity culminating with the trigger press.

Cross-Channel Correlation

Control subjects showed relatively high cross-channel correlations in all but the left-hemisphere derivations (C3' versus P3, with linked-ears reference) (Fig. 5). this finding, indicating a more focal mode of operation in the left hemisphere compared with the more diffuse, homogeneous right-hemisphere mode of control, agrees well with prevalent notions in the field of hemispheric asymmetry.[21] The respective correlation profiles of the TBI patients, howeveR, were in some cases different from the control profile. Figure 5 exemplifies cross-cortical correlation profiles associated with elbow extension in the TBI patients and in the control group. For left elbow extension, note the low correlation between the two parietal areas (P3 and P4) in subject AM as well as the low coherence (waveform correlation) within the right hemisphere of subject ST (left was the paretic side in the case of subject AM and the nonparetic side in subject ST's case). For right elbow extension, note the low coherence between the motor strips (C3'-C4') and within the right (paretic) hemisphere of subject DI. Of additional interest is the fact that an uncharacteristic profile may be associated with either the paretic or the nonparetic arm movements and may be disclosed in only a single condition.

Discussion

The amplitude analyses of the MRP records suggest that TBI patients show reduced premovement negativity associated with simple functional ballistic movements on the paretic side and, to a lesser degree, with movement on the nonparetic side. This reduced negativity, or, alternatively, the superimposed positive potential, may indicate a less adept motor control mechanism in TBI patients. The indication of functional impairments in the TBI patients is not in itself surprising. The ability to distinguish between different brain processes (eg, attentional versus motor) that may give rise to the reduction in the readiness potential, however, could be important to rehabilitation specialists.

The finding of reduced negativity, however, could simply reflect a ceiling effect attributable to the already elevated baseline (negative shift) occurring 1.5 seconds prior to the movement. This hypothesis has been eliminated by retesting two of the TBI patients (subjects TO and ST) and allowing for 3.5 seconds of pretrigger EEG activity to enter the averaging process. No early negative shift was apparent in these data. As in the initial data, both subjects exhibited reduced premovement negativity prior to the paretic hand movements and, to a lesser extent, prior to the nonparetic movements.

The unusually large potential shift between N2 and trigger activation for the paretic arm movement in the extension condition (Fig. 3) requires an explanation. This finding may be related to the nature of the task (ie, extending the paretic arm required more effort in the TBI patients than in the control subjects). Typically, closed head injury is associated with increased reflex activity of the paretic elbow brought about by a removal of the otherwise prominent cortical inhibition over other pathways. Elbow flexion is a component of the antigravity reflex set governed by the extra-pyramidal pathways.[22] Extending the elbow, therefore, would require greater physical or mental effort on the part of the TBI patients. Such effort, which most likely preceded the distal EMG activity, could have led to the TBI patients' motorlike cortical activity.

Reduced Negativity Versus

Superimposed Positivity

According to Karrer and coinvestigators,[14,15] reduced negativity prior to movement points to a less adept level of motor control. They have suggested that inhibitory processes secondary to motor inefficiency result in slow positive potentials, which are superimposed on the original negative shift. As argued earlier, the readiness potential may primarily reflect attentional processes. Furthermore, positive deflections in surface polarity could represent functional brain processes associated with the ongoing movement, whether such positivity originates in cortical relaxation (as suggested by Deecke et al[10]) or in deep subcortical activity. Either attentional impairments or reorganization (ie, a shift from cortical to subcortical focus) of movement control, therefore, could lead to the less negative/more positive slow premovement activity apparent in the TBI patients' data. In order to clarify these issues, two of the left-hemiparetic TBI patients and one control subject were tested using a contingent negative variation (CNV) paradigm.

Contingent Negative Variation

The CNV brain wave is typically recorded over the subject's skull during a fixed fore-period interval of a simple reaction-time paradigm. Because the electrical activity during this interval reflects preparation for movement and has been shown to resemble the slow negative shift (RP) in MRP experiments, [23] the CNV technique was used to help determine whether the reduction in premovement negativity in the TBI patients' data was associated with the allocation of attention or with actual motor programming of the ensuing movement. Subjects were asked to squeeze the trigger (finger flexion) as soon as they saw the "go" signal. The imperative ("go") signal was a small red light flashing for half a second on a light bar placed 1 m in front of the subject. A shorter-duration visual warning signal (100 milliseconds) was flashed 4 seconds prior to the "go" signal to prepare the subjects for the impending movement. The warning signal triggered an averaging process; 5-second intervals of EEG activity were captured and averaged off-line to allow for elimination of reactions that were too early or too late. The experimental apparatus was otherwise similar to the apparatus used in the MRP study. Figure 6 displays CNV waves associated with triggered reactions in two BTI patients and in one control subject.

In view of the electrical brain activity following the warning signal (Fig. 6), it is apparent that both TBI patients were capable of generating slow negative shifts over both hemispheres of the brain. This negative activity, however, peaked about 1 second following the warning signal and, unlike in the control subject, slowly decreased toward the end of the interstimulus interval (ISI), culminating in a small phasic negative potential just before the "go" signal. This negative peak, perhaps an expectancy potential, had a higher amplitude over the right (paretic) hemisphere (P4) in right (nonparetic) hand movements (subject TO). Hence, the premovement negative potential commonly associated with normal brain function can, under certain circumstances, indicate abnormal brain processes. Similarly, pronounced right parietal negativity was associated with left (paretic) hand movement (subject TO). This parietal activity peaked around the middle of the ISI and sustained its negative polarity for the remainder of the ISI. Such intensive parietal activity, which was not apparent in the control subject, may represent an abnormal elevation of the patient's attentional level as a result of the brain insult. In this specific case, the parietal association cortex in the paretic hemisphere of subject TO may compensate for functional flaws in the motor mechanisms and required additional attention to monitor the motor preparation.

Following the imperative ("go") signal, there were larger positive deflections over the motor strip of the TBI patients when performing the movement with the paretic hand (Fig. 6). The extensive phasic positivity preceding the motor wave may provide a clue to the reduced negativity observed in the TBI patients. Perhaps the same functional processes giving rise to the phasic positive deflection in triggered reactions are more widespread prior to voluntary movement and therefore reduce the underlying sustained premovement negative potential.

The MRP and CNV data together suggest that both attentional disorders and altered functional organization led to the reduced premotor negativity in the TBI patients. Although TBI patients seemingly are capable of generating negative potentials comparable to those of control subjects, the attentional mechanisms subserving the motor act in this patient population may often be less efficient and out of synchrony with this act. On the other hand, superimposed sustained or phasic positive potentials may re flect a different focus of control processes. Such positive potentials do not necessarily reflect defective or less-effective control mechanisms. They may simply indicate that the mode of control has shifted to a more automatic one and thus requries less attention. Such positive controls may occur as a result of brain injury, but could also reflect skill level or relative development stage.

The CNV records of one TBI patient (subject ST) may also offer an interpretation of the slow positive deflection over the right hemisphere during the RP interval (Fig. 4). The negative elevation, in this case, was discontinued a few hundred milliseconds prior to the motor act (paretic finger flexion). Similar deflection was apparent in the late CNV wave of this patient (Fig. 6), suggesting that attentional disorders specific to motor preparation were responsible for the apparent inability to sustain the negative shift.

Potential Usefulness of

MRP Traces to Clinical

Treatment Programs

Analyses of the individual MRP traces of the TBI patients revealed unusual brain potentials associated with specific electrode configurations as well as relatively low synchrony (correlation) between certain electrode derivations. This type of information, relating to the shape, location, and movement condition associated with pathological MRPs, may hold promise for noninvasive functional assessment of patients with closed head injury. For example, the posttrigger right parietal negativity recorded in right elbow extension (subject DI) resembles the directed attention potential (DAP) recorded by Deecke et al. [10] This parietal potential, believed to reflect special sensory preparation for an upcoming event, does not normally preced (or follow) simple ballistic movements such as those used in the present paradigm. Because the unusual potentials recorded over the right parietal hemisphere occur at the time when sensory information associated with the termination of the task is normally processed, they may indicate a problem concerning the integration of feedback information.

The physical therapist working with subject DI confirmed that the focus of the patient's deficits appeared to be sensory rather than motor in nature. If abnormal brain potentials were recorded over the motor strip prior to the movement, however, it would have suggested a functional difficulty associated with the motor programming of the ensuing movement, and treatment procedures would have been modified accordingly.

Finger movements have often been used to investigate preparatory cortical processes in humans. Fingers, however, have a large cortical representation, and use of the finger may differ in the nature of cerebral control processes from more proximal movements such as elbow flexion or extension. The latter movements most likely require more extensive subcortical contribution and thus may be sensitive to functional deficiencies that can go unnoticed if only finger movement is tested.

The data described in this article, although based on a small number of subjects, suggest that important information can be derived from brain potentials associated with simple movements. Careful observation of individual records and comparison of different techniques (ie, CNV and MRP) as well as of different analysis methods (eg, cross-cortical correlation) can help create individual profiles, or "signatures," which might explain some of the functional abnormalities apparent in pathological movement. Such profiles can also be derived from movements involving the nonparetic extremity, which makes the MRP technique a practical tool capable of assessing patients with severe hemiplegia as well.

The field of MRPs is young. Successful application of more complicated movement paradigms to clinical research will no doubt increase the diagnostic potential of this tool. Recent MRP investigation in healthy subjects [24] demonstrated brain potentials that corresponded to the kinematics and kinetics of the ongoing movement. Such "real" MRPs, along with pre-movement brain activity and potentials associated with preparation for movement (eg, CNV), should together comprise a promising, noninvasive assessment battery that will help the clinician determine functional abnormalities in pathological movements and plan more effective rehabilitation programs.

Knowledge of the nature of the functional deficit that leads to the movement impairment, whether it is primarily motor, primarily sensory, or attirubtable to behavior-specific attentional processes, is important for designing appropriate exercises and for choosing the most advantageous feedback mode in the rehabilitation regimen. Information regarding deficiencies in controlling the level of force or velocity in a given movement [24] could similarly be useful. For example, to remediate attentional impairments, a patient might practice tasks involving timing as well as augmented feedback. Difficulties of an efferent nature would call for more emphasis on rehabilitation of the neuromuscular system, such as can be done in physical training. Brain potentials associated with movements do not yet give us precise and reliable information. Yet, as this article has argued, the possible usefulness of this information is promising.

Conclusions

The TBI patients in this article showed reduced MRPs, compared with the control subjects, when performing the movement with their nonparetic arm. Very little MRP activity was associated with paretic arm movements. Subsequent testing of preparatory processes associated with a warned simple reaction-time paradigm (CNV) in two of these patients suggested that two different processes most likely contributed to the reduced movement-related slow negative potential in the TBI patients. Reduced negativity may be associated with defective attentional processes specific to the preparation and execution of the movement, or it may reflect functional reorganization of the motor system such as a relatively reduced cortical focus. Furthermore, a comparison between the MRP and CNV techniques allowed the interpretation of individual MRP traces and enabled discrimination between the attentional and motor processes. It is suggested that several measures of brain potentials associated with the preparation and execution of movement can together comprise a noninvasive test battery that will provide the rehabilitation worker with important information regarding the underlying functional abnomalities that lead to movement impairments.

(*1) Nicolet Biomedical Instruments, 5225-4 Verona Rd, PO Box 4287, Madison, WI 53711-0287.

References

[1] Vaughan HG, Costa LD, Ritter W. Topography of the human motor potential. Electroencephalogr Clin Neurophysiol. 1968;25:1-10.

[2] Kutas M, Donchin E. The effect of handedness, of responding hand, and of response force on the contralateral dominance of the readiness potential. In: Desmedt JE, ed. Attention, Voluntary Contraction and Event-Related Cerebral Potentials. Basel, Switzerland: S Karger AG, Medical and Scientific Publishers; 1977;1:189-210.

[3] Jergelova M. Distribution of the cortical potentials associated with voluntary finger movements in man. Activitas Nervosa Superior. 1980;22:233-240.

[4] Shibasaki H, Barrett G, Halliday AM, Halliday E. Scalp topography of movement-related cortical potentials. Prog Brain Res. 1980;54:237-242.

[5] Deecke L, Kornhuber HH, Lang W, et al. Timing function of the frontal cortex in sequential motor and learning tasks. Human Neurobiology. 1985;4:143-154.

[6] Kornhuber HH, Deecke L. Hirnpotentialanderungen beim Menshem vor und nach Willkurbewegungen, dargestellt mit Magnetbandspeicherung und Ruckwartsanalyse. Cited by: Deece L, Bashore T, Brunia CHM, et al. Movement-associated potentials and motor control: report of the EPIC VI Motor Panel. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:398-428.

[7] Kornhuber HH, Deecke L. Hirnopotentialanderungen bei Willkurbewegungen und passiven Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Cited by: Deecke L, Bashore T, Brunia CHM, et al. Movement-associated potentials and motor control: report of the EPIC VI Motor Panel. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:398-428.

[8] Kristeva R. Bereitschaftspotential of pianists. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:477-483.

[9] Goldberg G. Supplementary motor area structure and function: review and hypotheses. Behavioral and Brain Sciences. 1985;8;567-616.

[10] Deecke L, Bashore T, Brunia CHM, et al. Movement-associated potentials and motor control: report of the EPIC VI Motor Panel. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:398-428.

[11] Brunia CHM, Vingerhoets AJJM. CNV and EMG preceding a plantar flexion of the foot. Biol Psychol. 1980;11:181-191.

[12] Brunia CHM, Vingerhoets AJJM: Opposite hemisphere differences in movement-related potentials preceding foot and finger flexions. Biol Psychol. 1981;13:261-269.

[13] Kristeva R, Deecke L. Cerebral potentials preceding right and left unilateral and bilateral finger movements in sinistrals. In: Kornhuber HH, Deecke L, eds. Motivation: Motor and Sensory Processes of the Brain. Amsterdam, the Netherlands: Elsevier Science Publishers BV; 1980:748-754.

[14] Karrer R, Ivins J. Steady potentials accompanying perception and response in mentally retarded and normal children. In: Karrer R, ed. Developmental Psychophisiology of Mental Retardation. Springfield, Ill: Charles C Thomas, Publisher; 1976:361-417.

[15] Chisholm RC, Karrer R. Movement-related potentials and control of associated movements. Int J Neurosci. 1988;42:131-148.

[16] Bashore T. Movement-related potentials and handwriting posture. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:415-418.

[17] Grunewald-Zuberbier E, Grunewald G. Event-related brain potentials during voluntary ramp movements in aiming tasks. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:406-411.

[18] Shibasaki H, Barrett G, Neshige R, et al. Volitional movement is not preceded by cortical slow negativity in cerebellar dentate lesion in man. Brain Res. 1986;368:361-365.

[19] Deecke L, Englitz HG, Kornhuber HH, Schmitt G. Cerebral potentials preceding voluntary movement in patients with bilateral or unilateral Parkinson akinesia. In: Desmedt JE, ed. Attention, Voluntary Contraction and Event-Related Cerebral Potentials. Basel, Switzerland: S Karger AG, Medical and Scientific Publishers; 1977;1:151-163.

[20] Barrett G, Shibasaki H, Neshige R. Cortical potential shifts preceding voluntary movement are normal in parkinsonism. Electroencephalogr Clin Neurophysiol. 1987;63:340-348.

[21] Semmes J. Hemispheric specialization: a possible clue to mechanism. Neuropsychologia. 1968;6:11-26.

[22] Ghez C. Introduction to the motor systems. In: Kandel ER, Schwartz JH, eds. Principles of Neural Science. New York, NY: Elsevier Science Publishing Co Inc; 1981:271-284.

[23] Brunia CHM, Van Den Bosch WEJ: The influence of response side on the readiness potential prior to finger and foot movements: a preliminary report. In: Karrer R, Cohen J, Tueting P, eds. Brain and information: event-related potentials. Ann NY Acad Sci. 1984;425:434-438.

[24] York DH, Fausek D, Spagnolia T, Schafe M. Characterization of movement-related variables obtained from scalp-recorded cerebral activity in man. Society for Neuroscience Abstracts. 1989;15:397. Abstract.

A Nativ, PhD, was Research Associate, NeuroMuscular Retraining Clinic, Department of Rehabilitation Medicine, 2710 Marshall Ct, Madison, WI 53705, when this research was conducted. Dr Nativ is currently with the Department of Kinesiology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1.

This research was supported by NIDRR Grant G0087C2008.