Electroencephalography and clinical Neurophysiology, 82 (1992) 469-476
469
© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0013-4649/92/$05.00
EEG91578
Frontal DC potentials in auditory selective attention Susanne Asenbaum, Wilfried Lang, Alexander Egkher, Gerald Lindinger and Lfider Deecke Neurological University Clinic, Vienna (Austria) (Accepted for publication: 26 December 1991)
Summary Selective dichotic listening during periods of 35 sec was associated with negative shifts of the cortical DC potential. Amplitudes of negative DC potentials had maxima in frontal, in particular, in anterior frontal records. The temporal pattern of negative DC potentials was different between the fronto-lateral records of the two hemispheres: in records from the right side, DC potentials declined during the 35 sec observation period, whereas they remained sustained in those of the left side. Different instructions ("attend left ear," "attend right," "attend both") and different levels of pitch separation between deviants and standards had no effects on frontal negative DC potential shifts, which are discussed in terms of higher order control of selective dichotic listening. Key words: Slow negative potentials; Negative DC shifts; Selective attention; Frontal lobes
If the recording of the EEG is carried out with an infinite time constant, i.e., with a DC amplifier, then "slow potentials" or "DC potentials" can be picked up. DC potentials which are associated with the execution of a task are termed "event-related DC potentials." They persist at a constant level of amplitude during the task or develop and decay with a long time constant. Surface-recorded DC potentials typically reflect the depolarization of apical dendrites of pyramidal cells in layers I and II (Caspers et al. 1980; Speckmann and Elger 1987; for review: Rockstroh et al. 1989). They are associated with a variety of cognitive (Jung et al. 1984; Johnson et al. 1987; M. Lang et al. 1987; W. Lang et al. 1988a; Peronnet and Farah 1989; Ruchkin et al. 1990; Uhl et al. 1990a,b, 1991; R6sler et al. 1991) and motor (Griinewald et al. 1979; W. Lang et al. 1983, 1988b, 1989, 1990; Brunia and Damen 1988) tasks. In summary, these studies suggest two generalizations. First, the topography of the event-related DC potentials is different depending on task qualities. Analysis of the topographies yields information about the cortical areas which contribute to the surface-recorded DC potentials (Lang et al. 1990; Lindinger et al. 1990). Secondly, the amplitude of event-related DC potentials reflects the amount of effort which is employed in the task (R6sler et al. 1991). The present study investigated scalp-recorded DC potentials during periods of focused auditory attention.
Correspondence to: Susanne Asenbaum, M.D., Neurological Clinic, University of Vienna, W~ihringer Giirtel 18-20, A-1090 Vienna (Austria).
Its aim was to contribute towards understanding the physiological basis of attentional control. Recently, Hansen and Hillyard (1988) observed a slow negative potential shift arising at the onset of a period of focused attention. Improvements of recording techniques as employed in the present study made it possible to investigate the course of DC potentials across long periods of time (35 sec). Focused auditory attention was required in a selective dichotic listening situation. Experimental variations included the side of the designated ear (experiments I and II) and pitch separability between deviant and standard stimuli (experiment liD.
Methods
Three different experiments (I, II, III) were carried out. First, the common structure of all experiments will be described, then the differences.
Paradigm, stimuli A selective dichotic listening paradigm was used. Subjects initiated stimulus presentation by bimanual button pressing. Then, after a period of 1 sec, a small light spot informed the subject about the subsequent condition. After another period of 3 sec presentation of stimuli started and lasted for 35 sec. The stimuli (deviants, standards; intensity 72 dB SPL, duration 50 msec including 5 msec rise-time, 5 msec fall-time) were presented via earphones at interstimulus intervals (ISI) which randomly varied between 300 and 500 msec. The probability for standards was
470
S. ASENBAUM
0.75 and that of the deviants 0.25. Subjects had to press buttons bimanually with the two index fingers when detecting the deviant tone among the standards in a designated ear and to ignore all the input to the other ear.
Procedure Subjects were instructed to look at a focal point straight ahead and to prevent eye movements during the recording period which started 11 sec prior to the voluntary initiation of the task. After the 35 sec duration stimulus block subjects were allowed to rest. A minimum resting period of at least 30 sec was ensured because the computer rejected any premature starting signals initiated by the subjects. At least 30 artifact-free trials were performed for each of the two conditions. The sequence of the two conditions was randomized.
1
Data analysis Four data segments of 2 sec each were selected for further analysis (Fig. 1): 3-5, 10-12, 17-19 and 24-26 sec after the onset of the stimulus block at t = 4 sec. Mean amplitudes of DC potentials were calculated within those data segments. First, effects of condition
3
AL
4
I I I I Fpz
q,.,~
F7
~,-- -
F3
~,. .... - ' - ~ ~-"~
F4
" J
-
F8
~ ....
CZ
......
\"~
....
Y
T5 P3
Data acquisition The methods for reducing skin potential and for stabilizing electrode potentials as described by Bauer et al. (1989) were used in the present study. Nonpolarizable A g / A g C I electrodes were connected to the recording sites via salt bridges (silicon rubber tubes filled with electrode gel). Drifts of electrode potentials were lower than 3/zV within the analysis period which lasted 50 sec. Scratching of recording sites reduced electrode impedance to less than 1 kS2. The E E G was recorded in Fpz, F7, F8, F3, F4, Cz, P3, P4, T5 and T6 with linked ears serving for reference. Horizontal and vertical eye movements were controlled with two channels. The frequency band of amplification ranged between DC and 200 Hz (upper cut-off frequency). Resetting of the zero was done manually whenever DC potentials exceeded a range of + 100/zV at the input. Because of the low drifts resetting rarely had to be done. Data were digitized with a sampling rate of 200 Hz using a 12-bit analog to digital converter and stored on hard disc of a MicroVAX II. In an off-line procedure, each trial was controlled for artifacts caused by eye, head or orofacial movements. Criteria for rejecting a trial because of an artifact were (1) visible signs of electromyographic activity, (2) slow eye movements exceeding 40 /~V in the EOG, (3) drifts exceeding more than 40 ~V (comparing DC potentials of the first and the last second of the period of analysis). The baseline for measurement of DC shifts was taken from the first 8 sec of the analysis period.
2
ET
P4 T6
EOGv
IV --
__
;~-:"'~'~'~'.~:
I
2( pV
V
i,- ~ , ~
~
"
....
.
......
..
.
..--
.
...
.~ . . . . .
10 s
.
EOG h 0 4attend right teft .......... Fig. 1. Grand average of the DC potentials acquired during experiment I (attend right, dashed line, versus attend left, dotted line) across 12 subjects. At t = 0 voluntary task initiation occurred by bimanual button pressing and at t = 4 the auditory discrimination task started. 1, 2, 3 and 4 indicate the 4 data segments during which the mean amplitudes have been calculated (negativityup).
(COND) and stage of information processing (STAGE; 4 data segments) were tested by Multiple Analysis of Variance (MANOVA). This was done selectively for each electrode position. Then, hemispheric asymmetries were evaluated by introducing the within-subject factor SIDE which compared pairs of corresponding records (F7 vs. F8; F3 vs. F4, T5 vs. T6, P3 vs. P4).
Experiment I Twelve subjects (7 male, 5 female) ranging in age from 20 to 28 years (mean 24) participated in this experiment. All subjects were paid and were naive to the experimental tasks. Subjects were instructed to respond to deviant stimuli appearing either in the right ear (condition AR, "attend right") or the left ear (AL). The condition was announced by the light spot which appeared 1 sec after the voluntary initiation of the task (red light: AR; green light: AL). Deviants (1200 Hz) and standards (1000 Hz) were easy to separate. Within-subject factors for Multiple Analysis of Variance were COND (2 levels) and S T A G E (4 levels).
DC SHIFTS A N D A T T E N T I O N
471
Experiment II
Experiment 111
Ten subjects (6 males, 4 females) ranging in age from 20 to 26 years (mean 23) participated in this experiment. All subjects were naive. Only one condition was investigated which required subjects to press both buttons simultanously after the appearence of a deviant tone in either ear (left or right; condition AB, "attend both").
Ten subjects (7 males, 3 females; mean age: 24) participated in the third experiment. Two conditions were employed which differed by the frequency of the deviant tone. In an easy discrimination condition (ET, "easy task", announced by the red light) deviant tones of 1200 Hz were presented which could easily be distinguished from the standard tones (1000 Hz). In a
TABLE I Fpz
F7
F3
F4
F8
Cz
T5
P3
P4
T6
1.06
0.12
0.07
0.71
2.52
3.10
0.86
1.10
2.34
1.72
1.28
0.52
5.93 0.002 6.60 0.001
7.38 0.001 3.74 0.020
1.32
2.03
1.01
2.03
14.48 < 0.0001 5.10 0.005
2.33
3.02 0.04
3.53 0.025 2.53
2.48
7.77 < 0.001
9.53 < 0.0001
2.26
Experiment l Cond
dr l, 11 F =
Stage
dr3, 33
Condxstage
df3,33
Side
P: F = P: F = P:
df 1,11 F=
Side x stage
df 3, 33
Sidexcond
df 1, 11
Side x stage x cond
df 3, 33
P: F = P: F = P: F = P:
F7/8
F3/4
P3/4
T5/6
4.25
2.83
0.04
2.43
12.18 < 0.0001 4.63
3.91 ~017 1.32
2.11
0.37
0.34
0.26
2.38
3.66 0.022
1.20
1.32
Experiment H
Stage
df 3, 27 F =
Fpz
F7
F3
F4
F8
Cz
T5
P3
P4
T6
0.69
1.13
2.06
5.30 0.005
3.45 0.031
0.89
1.10
0.31
0.10
1.09
P:
Side
df 1, 9
Side x stage
df 3, 27
F = P: F = P:
F7/8
F3/4
P3/4
T5/6
0.18
1.00
0.01
1.15
3.78 0.022
4.63 0.010
0.38
1.15
Experiment III
Cond
df l, l l F = P: 3, 33 F = P: 3, 33 F = P:
Stage
df
Condxstage
df
Side
df 1,11 F =
Side x s t a g e
df 3, 33 F =
Fpz
F7
F3
F4
F8
Cz
T5
P3
P4
T6
0.05
0.29
0.50
0.00
3.54
2.26
0.07
0.69
1.35
0.00
0.21
0.54
4.96 0.007 0.52
0.36
0.28
1.37
2.61
3.48 0.03 2.99 0.048
1.30
0.52
13.41 < 0.0001 3.27 0.037
0.13
2.05
7.17 0.001 0.36
3.60 0.026
2.46
F7/8
F3/4
P3/4
T5/6
1.22
1.99
0.07
4.98 0.007 3.18
43.88 < 0.0001 9.18 < 0.0001 0.43
1.09
0.15
3.20
0.02
0.10
2.97
0.44
0.22
P:
Sidexcond
df 1, 11
Side x stage x cond
df 3, 33
P: F = P: F = P:
472
S. A S E N B A U M ET AL.
I
more difficult discrimination task (DT, green light) deviant tones were 1040 Hz. The instruction was to respond to deviant tones appearing in the right ear.
Fpz
Experiment I Event-related DC potentials.
F7
-'-- • . - v ~ " % "
F3
,~-- - -
Fpz
3
I III
Results
The voluntary initiation of the trial at t = 0 sec is preceded by a slow negative shift of the DC potential which starts rather early in the anterior frontal record (Fpz) and has its maximum there (Fig. 1). Since this potential shift precedes a voluntary movement it is termed Bereitschaftspotential (BP; Kornhuber and Deecke 1965). The period of selective dichotic listening (t = 4-39 sec) is associated with negative DC potentials which have their maxima in frontal records (Fpz, F7, F8, F3, F4). These negative DC potentials develop within a period of about 2 sec after the onset of the auditory task. The instructions, either to attend to the deviants in the right ear (condition AR) or in the left ear (condition AL) had no effects on DC potentials (Table I). In frontal records (F3, F4, F8) and Cz negative DC potentials decline across the period of 35 sec (significant effects of STAGE; Table I and Fig. 1). This
2
F4
~,- . . . .
- --,-~'J
T5
---
P3
~-
P4
T6
~
""
,-
-- - ~ ' - "
~--
V "~
- ~ " ~
[
2( IJV los
EOISv EOGh 0 4offend both Fig. 3. Grand average of the DC potentials acquired during experiment II (attend both) across 10 subjects. At t = 0 voluntary task initiation occurred by bimanual button pressing and at t = 4 the auditory discrimination task started. 1, 2, 3 and 4 indicate the 4 data segments during which the mean amplitudes have been calculated (negativity up).
F7
decline is larger in right fronto-lateral records than in corresponding left ones (significant interaction SIDE by STAGE for F 3 / F 4 and F7/F8; Table I). Evoked potentials. Evoked potentials of standards and deviants are different (Fig. 2) and, thus, indicate that an actual selective attention effect was present.
F8
Experiment II
T5
T6
..
....
-. . . . . . . . . . .
-...
I
100 ms
Fig. 2. Grand average of evoked potentials (1000 Hz standard tone, dashed line, versus 1200 Hz deviant tone, dotted line) during condition A R (attend right) across 12 subjects (negativity up).
The aim of this experiment was to test frontal lobe asymmetry in a selective dichotic listening task which required subjects to respond to deviants on either side (left and right ear; condition AB, "attend both"). Again, the voluntary initiation of the trial at t = 0 sec is preceded by a slow negative DC potential shift with maxima of amplitudes in frontal records (Fig. 3). Focused auditory attention during the period of dichotic listening is associated with negative DC potentials with a maximum in the anterior frontal record. These performance-related DC potentials significantly decline in F4 and F8 (significant effect of STAGE, Table I). This decline of amplitudes in fronto-lateral records is significantly larger in the right hemisphere
473
DC SHIFTS AND ATTENTION
1
2
3
t~
II: the decline of negative DC potentials is larger on the right than the left hemisphere (Table I and Fig. 4).
I I I I F p z
.
-
Discussion
-. ~
F7
..-'"
F3
-'-=
F4
.
.
.
...........,
.~..
.......
Cz
T
~
-
5
....
-
-
~
le
.
-
.-
.....
". . . . . . . . .
2q [~V T6
--
10 s
EOGv EOGh ~ . . . . 0
4
ectsy difficuti-
............
Fig. 4. Grand average of the DC potentials acquired during experiment III (easy condition, dashed line, versus difficult condition, dotted line) across 10 subjects. At t = 0 voluntary task initiation occurred by bimanual button pressing and at t = 4 the auditory discrimination task started. 1, 2, 3 and 4 indicates the 4 data segments during which the mean amplitudes have been calculated (negativity up).
than in the left (significant interaction SIDE x STAGE, Table I).
Experiment 111 The aim of the third experiment was to investigate effects of discrimination difficulty between deviants and standards on performance-related DC potentials. Two conditions were compared, one with easy pitch separability (ET, "easy task"), the other with difficult pitch separability (DT). As in the other two experiments a Bereitschaftspotential with a maximum of amplitude in the anterior frontal record preceded the voluntary initiation of the task (Fig. 4). The topography of DC potentials during focused auditory attention is similar to that of the previous two experiments. There is no difference between the two conditions indicating that variations of discrimination difficulty between deviant and standard stimuli do not affect performancerelated negative DC potentials (Table I). Hemispheric asymmetries of DC potentials in fronto-lateral records are consistent with the findings in experiments I and
DC potentials preceding the task Slow negative shifts of the DC potential precede the initiation of the task. In the anterior frontal record these anticiptory DC potentials start early, about 3 sec prior to the voluntary movement which initiates the task. Anticipatory DC potentials are negative in all records and have maxima of amplitudes in frontal records, in particular anterior frontal (Fpz). The experimental paradigm offered subjects the possibility to anticipate the occurrence of the selective attention task by employing a self-initiated movement for starting the trial. It is now accepted that not only self-initiated, voluntary movements but also non-motor events are preceded by anticipatory DC potentials on condition that they are of behavioural relevance and predictable in time of occurrence (Grtinewald and Griinewald-Zuberbier 1983; W. Lang et al. 1984, 1988b; Ruchkin et al. 1986; Brunia 1988). Anticipatory DC potentials preceding voluntary movements have been termed Bereitschaftspotential (BP; Kornhuber and Deecke 1965) or readiness potential; those prior to non-motor events have been termed stimulus preceding negativity (SPN) by Brunia (1988). Thus, the present paradigm fulfils the criteria to elicit both types of anticipatory DC potentials, BP and SPN. The point at issue is how to disentangle their contributions to the slow shift of the DC potential which was found in the present experiments. This is possible in a qualitative way since the characteristics of the BP prior to bimanual finger movements is already established (Kristeva et al. 1979; Benecke et al. 1985; Lang et al. 1988c): BP starts in Cz and has its maximum amplitude there. In frontal records the BP is small, sometimes absent or even positive with linked ears for reference. Considering this known BP topography it is evident that the large negative DC potentials in frontal records prior to task onset cannot be accounted for by the BP but reflect brain activity in anticipation of the selective attention task and can be classified as SPN. Present data may contribute towards understanding the functional significance of SPN by showing that anticiaptory DC potentials have a similar topography to those during the performance of the task. Rockstroh et al. (1989) have emphasized the existence of taskspecific anticipatory brain activity as indicated by DC potentials. Furthermore, it seems to be possible to separate BP and SPN with the paradigm employed in the present study. Systematic variations of the tasks may help to understand brain processes associated with anticipatory behaviour.
474
DC potentials associated with focused attention A first result is that focused attention in the selective dichotic listening task is associated with negative DC potentials which have maxima of amplitudes in frontal records, in particular anterior frontal. The existence of negative DC potentials during focused attention has already been demonstrated by Hansen and Hillyard (1988)who used the term "standing negativity" in order to describe the phenomenon. On the basis of measured distributions and size, these authors gave evidence that the "standing negativity" cannot be attributed to a simple linear summation of temporally smeared and overlapped auditory evoked potentials (N100 and Nd). Present data confirm that conclusion: DC potentials in Fpz have a mean size of about 18 izV, smearing of evoked potentials could account for less than 2 IzV. On the basis of topography the difference between DC potentials and evoked potentials becomes evident as well. Negative DC potentials are about 3 times larger in amplitude in Fpz than those in Cz (Figs. 1 and 2) while N100 is of similar size in Fpz and Cz. Also in good agreement with the data of Hansen and Hillyard (1988) DC potentials develop at the beginning of the selective attention period and reach their maxima within about 1-2 sec. The present data contribute towards understanding the physiological basis of focused attention by showing that the pattern of DC potentials changes during the 35 sec period. There is a significant decrease of amplitude in fronto-lateral records, which is quite consistent in the 3 experiments (experiment I: F3, F4, F8, Cz; experiment II: F4, F8; experiment III: F3, F4, F8). In other records, anterior frontal, parietal and temporal, amplitudes of DC potentials remain stable across the period of performance. In fronto-lateral records the decay of the negative DC potential was significantly larger in records of the right as compared to those of the left hemisphere. The hemispheric differences indicate that the relative contribution of the two hemispheres changes during the period. Since we know that "paradoxical asymmetries," i.e., neural source in the right hemisphere and negative DC potential over the left hemisphere and vice versa, only exist in situations with a neural source in mesial parts of a hemisphere (e.g., in foot movements, cf., Boschert and Deecke 1986) it is possible to conclude that it is the activity of the right hemisphere which decreases to a larger extent than that of the left one. From a functional point of view it can be concluded that brain processes underlying focused selective attention are composed at least of two components: one being "tonic" and related to a more sustained cortical activity in the left hemisphere, the other being "phasic" and associated with the decrease of cortical activity in the right hemisphere. Present data substantiate the model of Tucker and Williamson (1984) which is based
s. ASENBAUMET AL. on behavioural and neuropharmacological data. They distinguished between tonic attention and phasic arousal and related them to the left and the right hemisphere, respectively. In the context of these hemispheric asymmetries it may be interesting to note that the selective attention effect, i.e., the additional negative component (Nd, negativity difference) in attended evoked potentials as compared to unattended ones, is asymmetrically distributed over the two hemispheres, with larger amplitudes on the right than the left (Okita et al. 1983). Hemispheric asymmetries during focused visual and tactile attention have also been found in measurements of the regional cerebral blood flow (rCBF; Risberg and Prohovnik 1983). The fact that the DC potentials did not vary with the direction of spatial attention deserves discussion. In the visual and the somatosensory modalities directing attention to the left or the right hemifield of vision or part of the body, respectively, was associated with hemispheric differences in slow potentials prior to the presentation of the target. These slow potentials were surface-negative over the hemisphere contralateral to the direction of attention (Harter et al. 1982, 1989; Harter and Aine 1984; W. Lang et al. 1984); their amplitudes were about 1 pN. The scalp topography varied depending on the modality, with a parieto-occipital maximum in the visual and a centro-parietal maximum in the somatosensory task (W. Lang et al. 1984). Two explanations may account for the fact that negative DC potentials associated with the direction of hemispatial attention have not been found in the present study: (1) hemispatial attention of the auditory modality may not be organized asymmetically between the two hemispheres; however, neuropsychological studies demonstrate that hemispatial auditory inattention is associated with lesions in the contralateral hemisphere (e.g., Heilman 1979). (2) Scherg and Von Cramon (1986) demonstrated that in patients having a unilateral lesion the activation of the remaining auditory cortex produces an N100 which is reduced but still present over the damaged side. The anatomy of the auditory cortex may be too complex to dectect small asymmetrical activations due to hemispatial attention. The possibility might exist of detecting them when applying a larger number of recording positions and methods for modeling intracranial electrical activity. DC potentials during focused attention have maxima of negativity in frontal records, in particular anterior frontal. On condition that the event-related activity of more extended neural sources in the cortex mainly results in radially oriented current flow into the scalp (Speckmann and Elger 1987) we are allowed to assume that the frontal lobes, in particular anterior parts of it, are involved in the control of focused attention. Negative DC potentials of similar magnitude in frontal and anterior frontal records were found in
DC SHIFTS AND ATTENTION
tasks on motor and cognitive learning which were assumed to involve mainly the frontal lobes (paired associate learning: W. Lang et al. 1988a; Uhl et al. 1990b, 1991; concept formation: M. Lang et al. 1987; visuomotor learning: W. Lang et al. 1983, 1988b; retrieval from short-term memory: R6sler et al. 1991). One experiment of the previously mentioned studies enforced selectivity of information processing by introducing proactive interferences into a paired associate learning task (Uhl et al. 1991). Interestingly, the conclusion of that study was that the resistance against proactive interference was related to a large and sustained negative DC potentials in the anterior frontal record (Fpz). The demands on the control of focused attention in the present study are imposed by the necessity to process selectively the relevant stimulus in the designated ear and to resist the interference that arises from similarities between standards and deviants in the attended ear as well as from the appearence of deviants in the attended and unattended ear. Furthermore, the designation of the attended ear varied from trial to trial. Based on these similarities between the study of Uhl et al. (1991) and the present one, the hypothesis is raised that the anterior frontal cortex is involved in more general aspects of the control of selective information processing. Supported by the Austrian "Fonds zur F6rderung der Wissenschaft" (FWF; P8215).
References Bauer, H., Korunka, Ch. and Leodolter, M. Technical requirements for high-quality scalp DC recordings. Electroenceph. clin. Neurophysiol., 1989, 72: 545-547. Benecke, W., Dick, J.P.R., Rothwell, J.C., Day, B.D. and Marsden, C.D. Increase of the Bcreitschaftspotential in simultaneous and sequential movements. Neurosci. Lett., 1985, 62: 347-352. Boschert, J. and Deecke, L. Cerebral potentials preceding voluntary toe, knee and hip movements and their vectors in human precentral gyrus. Brain Res., 1986, 376: 175-179. Brunia, C.H.M. Movement and stimulus preceding negativity. Biol. Psychol., 1988, 26: 165-178. Brunia, C.H.M. and Damen, E.J.P. Distribution of slow brain potentials related to motor preparation and stimulus anticipation in a time estimation task. Electroenceph. clin. Neurophysiol., 1988, 69: 234-243. Caspers, H., Speckmann, E.J. and Lehmenkiihler, A. Electrogenesis of cortical DC potentials. In: H.H. Kornhuber and L. Deecke (Eds.). Motivation, Motor and Sensory Process of the Brain: Electrical Potentials, Behavior and Clinical Use. Progress in Brain Research, Vol. 54. Elsevier, Amsterdam, 1980: 3-16. GriJnewald, G. and Griinewald-Zuberbier, E. Cerebral potentials during voluntary ramp movements in aiming tasks. In: A.W.K. Gaillard and W. Ritter (Eds.), Tutorials in ERP Research: Endogenous Components. North Holland Publ., Amsterdam, 1983: 39-46. Griinewald, G., Griinewald-Zuberbier, E., H6mberg, V. and Netz, J. Cerebral potentials during smooth goal-directed hand movements in right-handed and left-handed subjects. Pfliigers Arch., 1979, 381: 39-46.
475 Hansen, J.C. and Hillyard, S.A. Temporal dynamics of human auditory selective attention. Psychophysiology, 1988, 25: 316-329. Harter, M.R. and Aine, C. Brain mechanisms of visual selective attention. In: R. Parasuraman and D.-R. Davies (Eds.), Varieties of Attention. Academic Press, New York, 1984: 293-321. Harter, M.R., Aine, C. and Schroeder, C. Hemispheric differences in the neural processing of stimulus location and type: effects of selective attention on visual evoked potentials. Neuropsychologia, 1982, 20: 421-437. Harter, M.R., Miller, S.L., Price, N.J., LaLonde, M.E. and Keyes, A.L. Neural processes involved in directing attention. J. Cogn. Neurosci., 1989, 1: 223-237. Heilmann, K.M. Neglect and related disorders. In: K.M. Heilmann and E. Valenstein (Eds.), Clinical Neuropsychology. Oxford University Press, New York, 1979: 268-307. Johnson, Jr., R., Cox, C. and Fedio, P. Event-related potential. Evidence for individual differences in mental rotation task. In: R. Johnson, Jr., J.W. Rohrbaugh and R. Parasuraman (Eds.), Current Trends in Event-Related Potential Research. Electroenceph. clin. Neurophysiol., Suppl 40. Elsevier, Amsterdam, 1987: 191197. Jung, R., Altenmiiller, E. and Natsch, B. Zur Hemisph~irendominanz fiir Sprache und Rechnen: elektroenzephalographische Korrelate einer Linksdominanz bei Linksh~indern. Neuropsychologia, 1984, 22: 755-775. Kornhuber, H.H. and Deecke, L. Hirnpotential~inderungen bei Willkiirbewegungen und passiven Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pfliigers Arch., 1965, 284: 1-17. Kristeva, R., Keller, E., Deecke, L. and Kornhuber, H.H. Cerebral potentials preceding unilateral and simultaneous bilateral finger movements. Electroenceph. clin. Neurophysiol., 1979, 47: 229238. Lang, M., Lang, W., Uhl, F., Kornhuber, A., Deecke, L. and Kornhuber, H.H. Slow potential shifts indicating verbal cognitive learning in a concept formation task. Hum. Neurobiol., 1987, 6: 183-190. Lang, W., Lang, M., Kornhuber, A., Deecke, L. and Kornhuber, H.H. Human cerebral potentials and visuomotor learning. Pflfigers Arch., 1983, 399: 342-344. Lang, W., Lang, M., Heise, B., Deecke, L. and Kornhuber, H.H. Brain potentials related to voluntary hand tracking, motivation and attention. Hum. Neurobiol., 1984, 3: 235-240. Lang, W., Lang, M., Uhl, F., Kornhuber, A., Deecke, L. and Kornhuber, H.H. Left frontal lobe in verbal associative learning: a slow potential study, Exp. Brain Res., 1988a, 70: 99-108. Lang, W., Lang, M., Podreka, I., Steiner, M., Uhl, F., Suess, E., Miiller, C. and Deecke, L. DC potential shifts and regional blood flow reveal frontal cortex involvement in human visomotor learning. Exp. Brain Res., 1988b, 71: 353-364. Lang, W., Lang, M., Uhl, F., Koska, Ch., Kornhuber, A. and Deecke, L. Negative cortical DC shifts preceding and accompanying simultaneous and sequential finger movements. Exp. Brain Res., 1988c, 71: 579-587. Lang, W., Zilch, O., Koska, C., Lindinger, G. and Deecke, L. Negative cortical DC shifts preceding and accompanying simple and complex sequential movements. Exp. Brain Res., 1989, 74: 99-104. Lang, W., Obrig, H., Lindinger, G., Cheyne, D. and Deecke, L. Supplementary motor area activation while tapping bimanually different rhythms in musicians. Exp. Brain Res., 1990, 79: 504514. Lindinger, G., Lang, W., Obrig, H. and Deecke, L. Current source density analysis of scalp potentials - topographical analysis of movement related DC shifts. In: C.H.M. Brunia, A.W.K. Gaillard and A. Kok (Eds.), Psychological Brain Research, Vol. 1. Tilburg University Press, Tilburg, 1990: 142-146.
476 Okita, T., Konishi, K. and Inamori, R. Attention-related negative brain potentials for speech words and pure tones. Biol. Psychol., 1983, 16: 29-47. Peronnet, F. and Farah, M.J. Mental rotation: an event-related potential study with a validated mental rotation task. Brain Cogn., 1989, 9: 279-288. Risberg, J. and Prohovnik, I. Cortical processing of visual and tactile stimuli studied by non-invasive rCBF measurements. Hum. Neurobiol., 1983, 2: 5-10. Rockstroh, B., Elbert, Th., Canavan, A., Lutzenberger, W. and Birbaumer, N. Slow Cortical Potentials and Behaviour. Urban and Schwarzenberg, Baltimore, MD, 1989. R6sler, F., Heil, M. and Glowalla, U. Monitoring retrieval from long-term memory by slow event-related brain potentials. Psychophysiology, 1991, in press. Ruchkin, D.S., Sutton, S., Mahaffey, D. and Glaser, J. Terminal CNV in the absence of motor response. Electroenceph. clin. Neurophysiol., 1986, 63: 445-463. Ruchkin, D.S., Johnson, Jr., R., Canoune, H. and Ritter, W. Shortterm memory storage and retention: an event-related brain potential study. Electroenceph. clin. Neurophysiol., 1990, 76: 419439.
S. ASENBAUM ET AL. Scherg, M. and Von Cramon, D. Evoked dipole source potentials of the human auditory cortex. Electroenceph. clin. Neurophysiol., 1986, 65: 344-360. Speckmann, E.J. and Elger, C.E. Neurophysiological basis of the EEG and DC potentials. In: E. Niedermeyer and F. Lopes da Silva (Eds.), Electroencephalography, 2nd edn. Urban and Schwarzenberg, Baltimore, MD, 1987: 1-13. Tucker, D.M. and Williamson, P.A. Asymmetric neural control systems in human self-regulation. Psychol. Rev., 1984, 91: 185-215. Uhl, F., Goldenberg, G., Lang, W., Lindinger, G., Steiner, M. and Deecke, L. Cerebral correlates of imagining colours, faces and a map. I1. Negative cortical DC potentials. Neuropsychologia, 1990a, 28: 81-93. Uhl, F., Lang, W., Lindinger, G. and Deecke, L. Elaborative strategy in word pair learning - DC-potential correlates of differential frontal and temporal lobe involvement. Neuropsychologia, 1990b, 28: 707-717. Uhl, F., Franzen, P., Series, W., Lang, W., Lindinger, G. and Deecke, L. Anterior frontal cortex and the effect of proactive interference in paired associate learning: a DC potential study. J. Cogn. Neurosci., 1991, 2: 373-382.
469
© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0013-4649/92/$05.00
EEG91578
Frontal DC potentials in auditory selective attention Susanne Asenbaum, Wilfried Lang, Alexander Egkher, Gerald Lindinger and Lfider Deecke Neurological University Clinic, Vienna (Austria) (Accepted for publication: 26 December 1991)
Summary Selective dichotic listening during periods of 35 sec was associated with negative shifts of the cortical DC potential. Amplitudes of negative DC potentials had maxima in frontal, in particular, in anterior frontal records. The temporal pattern of negative DC potentials was different between the fronto-lateral records of the two hemispheres: in records from the right side, DC potentials declined during the 35 sec observation period, whereas they remained sustained in those of the left side. Different instructions ("attend left ear," "attend right," "attend both") and different levels of pitch separation between deviants and standards had no effects on frontal negative DC potential shifts, which are discussed in terms of higher order control of selective dichotic listening. Key words: Slow negative potentials; Negative DC shifts; Selective attention; Frontal lobes
If the recording of the EEG is carried out with an infinite time constant, i.e., with a DC amplifier, then "slow potentials" or "DC potentials" can be picked up. DC potentials which are associated with the execution of a task are termed "event-related DC potentials." They persist at a constant level of amplitude during the task or develop and decay with a long time constant. Surface-recorded DC potentials typically reflect the depolarization of apical dendrites of pyramidal cells in layers I and II (Caspers et al. 1980; Speckmann and Elger 1987; for review: Rockstroh et al. 1989). They are associated with a variety of cognitive (Jung et al. 1984; Johnson et al. 1987; M. Lang et al. 1987; W. Lang et al. 1988a; Peronnet and Farah 1989; Ruchkin et al. 1990; Uhl et al. 1990a,b, 1991; R6sler et al. 1991) and motor (Griinewald et al. 1979; W. Lang et al. 1983, 1988b, 1989, 1990; Brunia and Damen 1988) tasks. In summary, these studies suggest two generalizations. First, the topography of the event-related DC potentials is different depending on task qualities. Analysis of the topographies yields information about the cortical areas which contribute to the surface-recorded DC potentials (Lang et al. 1990; Lindinger et al. 1990). Secondly, the amplitude of event-related DC potentials reflects the amount of effort which is employed in the task (R6sler et al. 1991). The present study investigated scalp-recorded DC potentials during periods of focused auditory attention.
Correspondence to: Susanne Asenbaum, M.D., Neurological Clinic, University of Vienna, W~ihringer Giirtel 18-20, A-1090 Vienna (Austria).
Its aim was to contribute towards understanding the physiological basis of attentional control. Recently, Hansen and Hillyard (1988) observed a slow negative potential shift arising at the onset of a period of focused attention. Improvements of recording techniques as employed in the present study made it possible to investigate the course of DC potentials across long periods of time (35 sec). Focused auditory attention was required in a selective dichotic listening situation. Experimental variations included the side of the designated ear (experiments I and II) and pitch separability between deviant and standard stimuli (experiment liD.
Methods
Three different experiments (I, II, III) were carried out. First, the common structure of all experiments will be described, then the differences.
Paradigm, stimuli A selective dichotic listening paradigm was used. Subjects initiated stimulus presentation by bimanual button pressing. Then, after a period of 1 sec, a small light spot informed the subject about the subsequent condition. After another period of 3 sec presentation of stimuli started and lasted for 35 sec. The stimuli (deviants, standards; intensity 72 dB SPL, duration 50 msec including 5 msec rise-time, 5 msec fall-time) were presented via earphones at interstimulus intervals (ISI) which randomly varied between 300 and 500 msec. The probability for standards was
470
S. ASENBAUM
0.75 and that of the deviants 0.25. Subjects had to press buttons bimanually with the two index fingers when detecting the deviant tone among the standards in a designated ear and to ignore all the input to the other ear.
Procedure Subjects were instructed to look at a focal point straight ahead and to prevent eye movements during the recording period which started 11 sec prior to the voluntary initiation of the task. After the 35 sec duration stimulus block subjects were allowed to rest. A minimum resting period of at least 30 sec was ensured because the computer rejected any premature starting signals initiated by the subjects. At least 30 artifact-free trials were performed for each of the two conditions. The sequence of the two conditions was randomized.
1
Data analysis Four data segments of 2 sec each were selected for further analysis (Fig. 1): 3-5, 10-12, 17-19 and 24-26 sec after the onset of the stimulus block at t = 4 sec. Mean amplitudes of DC potentials were calculated within those data segments. First, effects of condition
3
AL
4
I I I I Fpz
q,.,~
F7
~,-- -
F3
~,. .... - ' - ~ ~-"~
F4
" J
-
F8
~ ....
CZ
......
\"~
....
Y
T5 P3
Data acquisition The methods for reducing skin potential and for stabilizing electrode potentials as described by Bauer et al. (1989) were used in the present study. Nonpolarizable A g / A g C I electrodes were connected to the recording sites via salt bridges (silicon rubber tubes filled with electrode gel). Drifts of electrode potentials were lower than 3/zV within the analysis period which lasted 50 sec. Scratching of recording sites reduced electrode impedance to less than 1 kS2. The E E G was recorded in Fpz, F7, F8, F3, F4, Cz, P3, P4, T5 and T6 with linked ears serving for reference. Horizontal and vertical eye movements were controlled with two channels. The frequency band of amplification ranged between DC and 200 Hz (upper cut-off frequency). Resetting of the zero was done manually whenever DC potentials exceeded a range of + 100/zV at the input. Because of the low drifts resetting rarely had to be done. Data were digitized with a sampling rate of 200 Hz using a 12-bit analog to digital converter and stored on hard disc of a MicroVAX II. In an off-line procedure, each trial was controlled for artifacts caused by eye, head or orofacial movements. Criteria for rejecting a trial because of an artifact were (1) visible signs of electromyographic activity, (2) slow eye movements exceeding 40 /~V in the EOG, (3) drifts exceeding more than 40 ~V (comparing DC potentials of the first and the last second of the period of analysis). The baseline for measurement of DC shifts was taken from the first 8 sec of the analysis period.
2
ET
P4 T6
EOGv
IV --
__
;~-:"'~'~'~'.~:
I
2( pV
V
i,- ~ , ~
~
"
....
.
......
..
.
..--
.
...
.~ . . . . .
10 s
.
EOG h 0 4attend right teft .......... Fig. 1. Grand average of the DC potentials acquired during experiment I (attend right, dashed line, versus attend left, dotted line) across 12 subjects. At t = 0 voluntary task initiation occurred by bimanual button pressing and at t = 4 the auditory discrimination task started. 1, 2, 3 and 4 indicate the 4 data segments during which the mean amplitudes have been calculated (negativityup).
(COND) and stage of information processing (STAGE; 4 data segments) were tested by Multiple Analysis of Variance (MANOVA). This was done selectively for each electrode position. Then, hemispheric asymmetries were evaluated by introducing the within-subject factor SIDE which compared pairs of corresponding records (F7 vs. F8; F3 vs. F4, T5 vs. T6, P3 vs. P4).
Experiment I Twelve subjects (7 male, 5 female) ranging in age from 20 to 28 years (mean 24) participated in this experiment. All subjects were paid and were naive to the experimental tasks. Subjects were instructed to respond to deviant stimuli appearing either in the right ear (condition AR, "attend right") or the left ear (AL). The condition was announced by the light spot which appeared 1 sec after the voluntary initiation of the task (red light: AR; green light: AL). Deviants (1200 Hz) and standards (1000 Hz) were easy to separate. Within-subject factors for Multiple Analysis of Variance were COND (2 levels) and S T A G E (4 levels).
DC SHIFTS A N D A T T E N T I O N
471
Experiment II
Experiment 111
Ten subjects (6 males, 4 females) ranging in age from 20 to 26 years (mean 23) participated in this experiment. All subjects were naive. Only one condition was investigated which required subjects to press both buttons simultanously after the appearence of a deviant tone in either ear (left or right; condition AB, "attend both").
Ten subjects (7 males, 3 females; mean age: 24) participated in the third experiment. Two conditions were employed which differed by the frequency of the deviant tone. In an easy discrimination condition (ET, "easy task", announced by the red light) deviant tones of 1200 Hz were presented which could easily be distinguished from the standard tones (1000 Hz). In a
TABLE I Fpz
F7
F3
F4
F8
Cz
T5
P3
P4
T6
1.06
0.12
0.07
0.71
2.52
3.10
0.86
1.10
2.34
1.72
1.28
0.52
5.93 0.002 6.60 0.001
7.38 0.001 3.74 0.020
1.32
2.03
1.01
2.03
14.48 < 0.0001 5.10 0.005
2.33
3.02 0.04
3.53 0.025 2.53
2.48
7.77 < 0.001
9.53 < 0.0001
2.26
Experiment l Cond
dr l, 11 F =
Stage
dr3, 33
Condxstage
df3,33
Side
P: F = P: F = P:
df 1,11 F=
Side x stage
df 3, 33
Sidexcond
df 1, 11
Side x stage x cond
df 3, 33
P: F = P: F = P: F = P:
F7/8
F3/4
P3/4
T5/6
4.25
2.83
0.04
2.43
12.18 < 0.0001 4.63
3.91 ~017 1.32
2.11
0.37
0.34
0.26
2.38
3.66 0.022
1.20
1.32
Experiment H
Stage
df 3, 27 F =
Fpz
F7
F3
F4
F8
Cz
T5
P3
P4
T6
0.69
1.13
2.06
5.30 0.005
3.45 0.031
0.89
1.10
0.31
0.10
1.09
P:
Side
df 1, 9
Side x stage
df 3, 27
F = P: F = P:
F7/8
F3/4
P3/4
T5/6
0.18
1.00
0.01
1.15
3.78 0.022
4.63 0.010
0.38
1.15
Experiment III
Cond
df l, l l F = P: 3, 33 F = P: 3, 33 F = P:
Stage
df
Condxstage
df
Side
df 1,11 F =
Side x s t a g e
df 3, 33 F =
Fpz
F7
F3
F4
F8
Cz
T5
P3
P4
T6
0.05
0.29
0.50
0.00
3.54
2.26
0.07
0.69
1.35
0.00
0.21
0.54
4.96 0.007 0.52
0.36
0.28
1.37
2.61
3.48 0.03 2.99 0.048
1.30
0.52
13.41 < 0.0001 3.27 0.037
0.13
2.05
7.17 0.001 0.36
3.60 0.026
2.46
F7/8
F3/4
P3/4
T5/6
1.22
1.99
0.07
4.98 0.007 3.18
43.88 < 0.0001 9.18 < 0.0001 0.43
1.09
0.15
3.20
0.02
0.10
2.97
0.44
0.22
P:
Sidexcond
df 1, 11
Side x stage x cond
df 3, 33
P: F = P: F = P:
472
S. A S E N B A U M ET AL.
I
more difficult discrimination task (DT, green light) deviant tones were 1040 Hz. The instruction was to respond to deviant tones appearing in the right ear.
Fpz
Experiment I Event-related DC potentials.
F7
-'-- • . - v ~ " % "
F3
,~-- - -
Fpz
3
I III
Results
The voluntary initiation of the trial at t = 0 sec is preceded by a slow negative shift of the DC potential which starts rather early in the anterior frontal record (Fpz) and has its maximum there (Fig. 1). Since this potential shift precedes a voluntary movement it is termed Bereitschaftspotential (BP; Kornhuber and Deecke 1965). The period of selective dichotic listening (t = 4-39 sec) is associated with negative DC potentials which have their maxima in frontal records (Fpz, F7, F8, F3, F4). These negative DC potentials develop within a period of about 2 sec after the onset of the auditory task. The instructions, either to attend to the deviants in the right ear (condition AR) or in the left ear (condition AL) had no effects on DC potentials (Table I). In frontal records (F3, F4, F8) and Cz negative DC potentials decline across the period of 35 sec (significant effects of STAGE; Table I and Fig. 1). This
2
F4
~,- . . . .
- --,-~'J
T5
---
P3
~-
P4
T6
~
""
,-
-- - ~ ' - "
~--
V "~
- ~ " ~
[
2( IJV los
EOISv EOGh 0 4offend both Fig. 3. Grand average of the DC potentials acquired during experiment II (attend both) across 10 subjects. At t = 0 voluntary task initiation occurred by bimanual button pressing and at t = 4 the auditory discrimination task started. 1, 2, 3 and 4 indicate the 4 data segments during which the mean amplitudes have been calculated (negativity up).
F7
decline is larger in right fronto-lateral records than in corresponding left ones (significant interaction SIDE by STAGE for F 3 / F 4 and F7/F8; Table I). Evoked potentials. Evoked potentials of standards and deviants are different (Fig. 2) and, thus, indicate that an actual selective attention effect was present.
F8
Experiment II
T5
T6
..
....
-. . . . . . . . . . .
-...
I
100 ms
Fig. 2. Grand average of evoked potentials (1000 Hz standard tone, dashed line, versus 1200 Hz deviant tone, dotted line) during condition A R (attend right) across 12 subjects (negativity up).
The aim of this experiment was to test frontal lobe asymmetry in a selective dichotic listening task which required subjects to respond to deviants on either side (left and right ear; condition AB, "attend both"). Again, the voluntary initiation of the trial at t = 0 sec is preceded by a slow negative DC potential shift with maxima of amplitudes in frontal records (Fig. 3). Focused auditory attention during the period of dichotic listening is associated with negative DC potentials with a maximum in the anterior frontal record. These performance-related DC potentials significantly decline in F4 and F8 (significant effect of STAGE, Table I). This decline of amplitudes in fronto-lateral records is significantly larger in the right hemisphere
473
DC SHIFTS AND ATTENTION
1
2
3
t~
II: the decline of negative DC potentials is larger on the right than the left hemisphere (Table I and Fig. 4).
I I I I F p z
.
-
Discussion
-. ~
F7
..-'"
F3
-'-=
F4
.
.
.
...........,
.~..
.......
Cz
T
~
-
5
....
-
-
~
le
.
-
.-
.....
". . . . . . . . .
2q [~V T6
--
10 s
EOGv EOGh ~ . . . . 0
4
ectsy difficuti-
............
Fig. 4. Grand average of the DC potentials acquired during experiment III (easy condition, dashed line, versus difficult condition, dotted line) across 10 subjects. At t = 0 voluntary task initiation occurred by bimanual button pressing and at t = 4 the auditory discrimination task started. 1, 2, 3 and 4 indicates the 4 data segments during which the mean amplitudes have been calculated (negativity up).
than in the left (significant interaction SIDE x STAGE, Table I).
Experiment 111 The aim of the third experiment was to investigate effects of discrimination difficulty between deviants and standards on performance-related DC potentials. Two conditions were compared, one with easy pitch separability (ET, "easy task"), the other with difficult pitch separability (DT). As in the other two experiments a Bereitschaftspotential with a maximum of amplitude in the anterior frontal record preceded the voluntary initiation of the task (Fig. 4). The topography of DC potentials during focused auditory attention is similar to that of the previous two experiments. There is no difference between the two conditions indicating that variations of discrimination difficulty between deviant and standard stimuli do not affect performancerelated negative DC potentials (Table I). Hemispheric asymmetries of DC potentials in fronto-lateral records are consistent with the findings in experiments I and
DC potentials preceding the task Slow negative shifts of the DC potential precede the initiation of the task. In the anterior frontal record these anticiptory DC potentials start early, about 3 sec prior to the voluntary movement which initiates the task. Anticipatory DC potentials are negative in all records and have maxima of amplitudes in frontal records, in particular anterior frontal (Fpz). The experimental paradigm offered subjects the possibility to anticipate the occurrence of the selective attention task by employing a self-initiated movement for starting the trial. It is now accepted that not only self-initiated, voluntary movements but also non-motor events are preceded by anticipatory DC potentials on condition that they are of behavioural relevance and predictable in time of occurrence (Grtinewald and Griinewald-Zuberbier 1983; W. Lang et al. 1984, 1988b; Ruchkin et al. 1986; Brunia 1988). Anticipatory DC potentials preceding voluntary movements have been termed Bereitschaftspotential (BP; Kornhuber and Deecke 1965) or readiness potential; those prior to non-motor events have been termed stimulus preceding negativity (SPN) by Brunia (1988). Thus, the present paradigm fulfils the criteria to elicit both types of anticipatory DC potentials, BP and SPN. The point at issue is how to disentangle their contributions to the slow shift of the DC potential which was found in the present experiments. This is possible in a qualitative way since the characteristics of the BP prior to bimanual finger movements is already established (Kristeva et al. 1979; Benecke et al. 1985; Lang et al. 1988c): BP starts in Cz and has its maximum amplitude there. In frontal records the BP is small, sometimes absent or even positive with linked ears for reference. Considering this known BP topography it is evident that the large negative DC potentials in frontal records prior to task onset cannot be accounted for by the BP but reflect brain activity in anticipation of the selective attention task and can be classified as SPN. Present data may contribute towards understanding the functional significance of SPN by showing that anticiaptory DC potentials have a similar topography to those during the performance of the task. Rockstroh et al. (1989) have emphasized the existence of taskspecific anticipatory brain activity as indicated by DC potentials. Furthermore, it seems to be possible to separate BP and SPN with the paradigm employed in the present study. Systematic variations of the tasks may help to understand brain processes associated with anticipatory behaviour.
474
DC potentials associated with focused attention A first result is that focused attention in the selective dichotic listening task is associated with negative DC potentials which have maxima of amplitudes in frontal records, in particular anterior frontal. The existence of negative DC potentials during focused attention has already been demonstrated by Hansen and Hillyard (1988)who used the term "standing negativity" in order to describe the phenomenon. On the basis of measured distributions and size, these authors gave evidence that the "standing negativity" cannot be attributed to a simple linear summation of temporally smeared and overlapped auditory evoked potentials (N100 and Nd). Present data confirm that conclusion: DC potentials in Fpz have a mean size of about 18 izV, smearing of evoked potentials could account for less than 2 IzV. On the basis of topography the difference between DC potentials and evoked potentials becomes evident as well. Negative DC potentials are about 3 times larger in amplitude in Fpz than those in Cz (Figs. 1 and 2) while N100 is of similar size in Fpz and Cz. Also in good agreement with the data of Hansen and Hillyard (1988) DC potentials develop at the beginning of the selective attention period and reach their maxima within about 1-2 sec. The present data contribute towards understanding the physiological basis of focused attention by showing that the pattern of DC potentials changes during the 35 sec period. There is a significant decrease of amplitude in fronto-lateral records, which is quite consistent in the 3 experiments (experiment I: F3, F4, F8, Cz; experiment II: F4, F8; experiment III: F3, F4, F8). In other records, anterior frontal, parietal and temporal, amplitudes of DC potentials remain stable across the period of performance. In fronto-lateral records the decay of the negative DC potential was significantly larger in records of the right as compared to those of the left hemisphere. The hemispheric differences indicate that the relative contribution of the two hemispheres changes during the period. Since we know that "paradoxical asymmetries," i.e., neural source in the right hemisphere and negative DC potential over the left hemisphere and vice versa, only exist in situations with a neural source in mesial parts of a hemisphere (e.g., in foot movements, cf., Boschert and Deecke 1986) it is possible to conclude that it is the activity of the right hemisphere which decreases to a larger extent than that of the left one. From a functional point of view it can be concluded that brain processes underlying focused selective attention are composed at least of two components: one being "tonic" and related to a more sustained cortical activity in the left hemisphere, the other being "phasic" and associated with the decrease of cortical activity in the right hemisphere. Present data substantiate the model of Tucker and Williamson (1984) which is based
s. ASENBAUMET AL. on behavioural and neuropharmacological data. They distinguished between tonic attention and phasic arousal and related them to the left and the right hemisphere, respectively. In the context of these hemispheric asymmetries it may be interesting to note that the selective attention effect, i.e., the additional negative component (Nd, negativity difference) in attended evoked potentials as compared to unattended ones, is asymmetrically distributed over the two hemispheres, with larger amplitudes on the right than the left (Okita et al. 1983). Hemispheric asymmetries during focused visual and tactile attention have also been found in measurements of the regional cerebral blood flow (rCBF; Risberg and Prohovnik 1983). The fact that the DC potentials did not vary with the direction of spatial attention deserves discussion. In the visual and the somatosensory modalities directing attention to the left or the right hemifield of vision or part of the body, respectively, was associated with hemispheric differences in slow potentials prior to the presentation of the target. These slow potentials were surface-negative over the hemisphere contralateral to the direction of attention (Harter et al. 1982, 1989; Harter and Aine 1984; W. Lang et al. 1984); their amplitudes were about 1 pN. The scalp topography varied depending on the modality, with a parieto-occipital maximum in the visual and a centro-parietal maximum in the somatosensory task (W. Lang et al. 1984). Two explanations may account for the fact that negative DC potentials associated with the direction of hemispatial attention have not been found in the present study: (1) hemispatial attention of the auditory modality may not be organized asymmetically between the two hemispheres; however, neuropsychological studies demonstrate that hemispatial auditory inattention is associated with lesions in the contralateral hemisphere (e.g., Heilman 1979). (2) Scherg and Von Cramon (1986) demonstrated that in patients having a unilateral lesion the activation of the remaining auditory cortex produces an N100 which is reduced but still present over the damaged side. The anatomy of the auditory cortex may be too complex to dectect small asymmetrical activations due to hemispatial attention. The possibility might exist of detecting them when applying a larger number of recording positions and methods for modeling intracranial electrical activity. DC potentials during focused attention have maxima of negativity in frontal records, in particular anterior frontal. On condition that the event-related activity of more extended neural sources in the cortex mainly results in radially oriented current flow into the scalp (Speckmann and Elger 1987) we are allowed to assume that the frontal lobes, in particular anterior parts of it, are involved in the control of focused attention. Negative DC potentials of similar magnitude in frontal and anterior frontal records were found in
DC SHIFTS AND ATTENTION
tasks on motor and cognitive learning which were assumed to involve mainly the frontal lobes (paired associate learning: W. Lang et al. 1988a; Uhl et al. 1990b, 1991; concept formation: M. Lang et al. 1987; visuomotor learning: W. Lang et al. 1983, 1988b; retrieval from short-term memory: R6sler et al. 1991). One experiment of the previously mentioned studies enforced selectivity of information processing by introducing proactive interferences into a paired associate learning task (Uhl et al. 1991). Interestingly, the conclusion of that study was that the resistance against proactive interference was related to a large and sustained negative DC potentials in the anterior frontal record (Fpz). The demands on the control of focused attention in the present study are imposed by the necessity to process selectively the relevant stimulus in the designated ear and to resist the interference that arises from similarities between standards and deviants in the attended ear as well as from the appearence of deviants in the attended and unattended ear. Furthermore, the designation of the attended ear varied from trial to trial. Based on these similarities between the study of Uhl et al. (1991) and the present one, the hypothesis is raised that the anterior frontal cortex is involved in more general aspects of the control of selective information processing. Supported by the Austrian "Fonds zur F6rderung der Wissenschaft" (FWF; P8215).
References Bauer, H., Korunka, Ch. and Leodolter, M. Technical requirements for high-quality scalp DC recordings. Electroenceph. clin. Neurophysiol., 1989, 72: 545-547. Benecke, W., Dick, J.P.R., Rothwell, J.C., Day, B.D. and Marsden, C.D. Increase of the Bcreitschaftspotential in simultaneous and sequential movements. Neurosci. Lett., 1985, 62: 347-352. Boschert, J. and Deecke, L. Cerebral potentials preceding voluntary toe, knee and hip movements and their vectors in human precentral gyrus. Brain Res., 1986, 376: 175-179. Brunia, C.H.M. Movement and stimulus preceding negativity. Biol. Psychol., 1988, 26: 165-178. Brunia, C.H.M. and Damen, E.J.P. Distribution of slow brain potentials related to motor preparation and stimulus anticipation in a time estimation task. Electroenceph. clin. Neurophysiol., 1988, 69: 234-243. Caspers, H., Speckmann, E.J. and Lehmenkiihler, A. Electrogenesis of cortical DC potentials. In: H.H. Kornhuber and L. Deecke (Eds.). Motivation, Motor and Sensory Process of the Brain: Electrical Potentials, Behavior and Clinical Use. Progress in Brain Research, Vol. 54. Elsevier, Amsterdam, 1980: 3-16. GriJnewald, G. and Griinewald-Zuberbier, E. Cerebral potentials during voluntary ramp movements in aiming tasks. In: A.W.K. Gaillard and W. Ritter (Eds.), Tutorials in ERP Research: Endogenous Components. North Holland Publ., Amsterdam, 1983: 39-46. Griinewald, G., Griinewald-Zuberbier, E., H6mberg, V. and Netz, J. Cerebral potentials during smooth goal-directed hand movements in right-handed and left-handed subjects. Pfliigers Arch., 1979, 381: 39-46.
475 Hansen, J.C. and Hillyard, S.A. Temporal dynamics of human auditory selective attention. Psychophysiology, 1988, 25: 316-329. Harter, M.R. and Aine, C. Brain mechanisms of visual selective attention. In: R. Parasuraman and D.-R. Davies (Eds.), Varieties of Attention. Academic Press, New York, 1984: 293-321. Harter, M.R., Aine, C. and Schroeder, C. Hemispheric differences in the neural processing of stimulus location and type: effects of selective attention on visual evoked potentials. Neuropsychologia, 1982, 20: 421-437. Harter, M.R., Miller, S.L., Price, N.J., LaLonde, M.E. and Keyes, A.L. Neural processes involved in directing attention. J. Cogn. Neurosci., 1989, 1: 223-237. Heilmann, K.M. Neglect and related disorders. In: K.M. Heilmann and E. Valenstein (Eds.), Clinical Neuropsychology. Oxford University Press, New York, 1979: 268-307. Johnson, Jr., R., Cox, C. and Fedio, P. Event-related potential. Evidence for individual differences in mental rotation task. In: R. Johnson, Jr., J.W. Rohrbaugh and R. Parasuraman (Eds.), Current Trends in Event-Related Potential Research. Electroenceph. clin. Neurophysiol., Suppl 40. Elsevier, Amsterdam, 1987: 191197. Jung, R., Altenmiiller, E. and Natsch, B. Zur Hemisph~irendominanz fiir Sprache und Rechnen: elektroenzephalographische Korrelate einer Linksdominanz bei Linksh~indern. Neuropsychologia, 1984, 22: 755-775. Kornhuber, H.H. and Deecke, L. Hirnpotential~inderungen bei Willkiirbewegungen und passiven Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pfliigers Arch., 1965, 284: 1-17. Kristeva, R., Keller, E., Deecke, L. and Kornhuber, H.H. Cerebral potentials preceding unilateral and simultaneous bilateral finger movements. Electroenceph. clin. Neurophysiol., 1979, 47: 229238. Lang, M., Lang, W., Uhl, F., Kornhuber, A., Deecke, L. and Kornhuber, H.H. Slow potential shifts indicating verbal cognitive learning in a concept formation task. Hum. Neurobiol., 1987, 6: 183-190. Lang, W., Lang, M., Kornhuber, A., Deecke, L. and Kornhuber, H.H. Human cerebral potentials and visuomotor learning. Pflfigers Arch., 1983, 399: 342-344. Lang, W., Lang, M., Heise, B., Deecke, L. and Kornhuber, H.H. Brain potentials related to voluntary hand tracking, motivation and attention. Hum. Neurobiol., 1984, 3: 235-240. Lang, W., Lang, M., Uhl, F., Kornhuber, A., Deecke, L. and Kornhuber, H.H. Left frontal lobe in verbal associative learning: a slow potential study, Exp. Brain Res., 1988a, 70: 99-108. Lang, W., Lang, M., Podreka, I., Steiner, M., Uhl, F., Suess, E., Miiller, C. and Deecke, L. DC potential shifts and regional blood flow reveal frontal cortex involvement in human visomotor learning. Exp. Brain Res., 1988b, 71: 353-364. Lang, W., Lang, M., Uhl, F., Koska, Ch., Kornhuber, A. and Deecke, L. Negative cortical DC shifts preceding and accompanying simultaneous and sequential finger movements. Exp. Brain Res., 1988c, 71: 579-587. Lang, W., Zilch, O., Koska, C., Lindinger, G. and Deecke, L. Negative cortical DC shifts preceding and accompanying simple and complex sequential movements. Exp. Brain Res., 1989, 74: 99-104. Lang, W., Obrig, H., Lindinger, G., Cheyne, D. and Deecke, L. Supplementary motor area activation while tapping bimanually different rhythms in musicians. Exp. Brain Res., 1990, 79: 504514. Lindinger, G., Lang, W., Obrig, H. and Deecke, L. Current source density analysis of scalp potentials - topographical analysis of movement related DC shifts. In: C.H.M. Brunia, A.W.K. Gaillard and A. Kok (Eds.), Psychological Brain Research, Vol. 1. Tilburg University Press, Tilburg, 1990: 142-146.
476 Okita, T., Konishi, K. and Inamori, R. Attention-related negative brain potentials for speech words and pure tones. Biol. Psychol., 1983, 16: 29-47. Peronnet, F. and Farah, M.J. Mental rotation: an event-related potential study with a validated mental rotation task. Brain Cogn., 1989, 9: 279-288. Risberg, J. and Prohovnik, I. Cortical processing of visual and tactile stimuli studied by non-invasive rCBF measurements. Hum. Neurobiol., 1983, 2: 5-10. Rockstroh, B., Elbert, Th., Canavan, A., Lutzenberger, W. and Birbaumer, N. Slow Cortical Potentials and Behaviour. Urban and Schwarzenberg, Baltimore, MD, 1989. R6sler, F., Heil, M. and Glowalla, U. Monitoring retrieval from long-term memory by slow event-related brain potentials. Psychophysiology, 1991, in press. Ruchkin, D.S., Sutton, S., Mahaffey, D. and Glaser, J. Terminal CNV in the absence of motor response. Electroenceph. clin. Neurophysiol., 1986, 63: 445-463. Ruchkin, D.S., Johnson, Jr., R., Canoune, H. and Ritter, W. Shortterm memory storage and retention: an event-related brain potential study. Electroenceph. clin. Neurophysiol., 1990, 76: 419439.
S. ASENBAUM ET AL. Scherg, M. and Von Cramon, D. Evoked dipole source potentials of the human auditory cortex. Electroenceph. clin. Neurophysiol., 1986, 65: 344-360. Speckmann, E.J. and Elger, C.E. Neurophysiological basis of the EEG and DC potentials. In: E. Niedermeyer and F. Lopes da Silva (Eds.), Electroencephalography, 2nd edn. Urban and Schwarzenberg, Baltimore, MD, 1987: 1-13. Tucker, D.M. and Williamson, P.A. Asymmetric neural control systems in human self-regulation. Psychol. Rev., 1984, 91: 185-215. Uhl, F., Goldenberg, G., Lang, W., Lindinger, G., Steiner, M. and Deecke, L. Cerebral correlates of imagining colours, faces and a map. I1. Negative cortical DC potentials. Neuropsychologia, 1990a, 28: 81-93. Uhl, F., Lang, W., Lindinger, G. and Deecke, L. Elaborative strategy in word pair learning - DC-potential correlates of differential frontal and temporal lobe involvement. Neuropsychologia, 1990b, 28: 707-717. Uhl, F., Franzen, P., Series, W., Lang, W., Lindinger, G. and Deecke, L. Anterior frontal cortex and the effect of proactive interference in paired associate learning: a DC potential study. J. Cogn. Neurosci., 1991, 2: 373-382.
