- Email: [email protected]
Evoked magnetic responses of the human auditory cortex to minor pitch changes: localization of the mismatch field
538
Electroenccphalography and clinical Neurophysiology, 84 (1992) 538-548 ~,~'~1992 Elsevier Scientific Publishers Ireland, Ltd. [1168-5597/92/$05.011
EVOPOT 925115
Evoked magnetic responses of the human auditory cortex to minor pitch changes: localization of the m i s m a t c h field * V. Cs6pe **, C. Pantev, M. Hoke, S. Hampson and B. Ross Institute ~ff"Experimet~tal Audiology, Unit'ersity of" Miinster, Miinster (Germany) (Accepted for publication: 23 July 1992)
Summary The neuromagnetic source Iocalizations of the auditory M 100 and the mismatch field (MMF) were studied using a large-array biomagnetometer. Standard tones of 1000 Hz and deviant tones of 11150 Hz were delivered with 90% and 10'~ probability, respectively. Wave forms of the derived MMF were computed by examining difference wave forms between the responses to the deviants and the responses to the standards preceding (D-P) and following (D-F) the deviants as well as to all remaining standards (D-A). The subset of standards preceding the deviants was used for a more realistic comparison with the set of deviants (having the same number of epochs and a similar signal-to-noise ratio), while the subset of standards following the deviants served to answer the question whether those standards also elicit an MMF. The MMF deflections were compared with each other, with the "'native" MMF occurring in response to the deviants, and with wave M 100. (The MMF as it appears in the unprocessed response to the deviants was termed "'native" for an easy distinction from the "derived" MMF.) Our results demonstrate a distinct MMF deflection, corresponding in latency to the simultaneously recorded fronto-central electrical MMN. Source analysis, using a single moving dipole model, showed the same spatial localization for the native MMF and for the different derived MMFs. The MMF source location turned out to be significantly anterior, medial and inferior relative to the sources of the M 1110. The present data also demonstrate that a minor frequency deviation may not activate measurably different MI00 generators, yet be sufficient It) trigger the nearby but spatially distinct mismatch generator. Key words: Auditory cortex; Auditory evoked magnetic field; Mismatch field; Derived mismatch field; Neuromagnetic source localization
A negalive component of the event-related brain potential (ERP), called the mismatch negativity ( M M N ) , d e s c r i b e d first by N~i~it~inen et al. (1978), o c c u r s in r e s p o n s e to a d e v i a n t s t i m u l u s p r e s e n t e d , i n an o t h e r w i s e h o m o g e n e o u s s e q u e n c e o f s t a n d a r d stimuli. T h e s e a u t h o r s i s o l a t e d t h e M M N f r o m w a v e N 2 d e s c r i b e d in m a n y p r e v i o u s s t u d i e s ( F o r d et al. 1976a,b; S i m s o n et al. 1977; S q u i r e s et al. 1977). L a t e r , N~i~it~inen a n d his c o - w o r k e r s (N~i~it~inen a n d G a i l l a r d 1983) sugg e s t e d t h a t w a v e N 2 is c o m p o s e d o f two c o m p o n e n t s , t h e N 2 b a n d M M N . W h i l e t h e M M N c a n be e l i c i t e d e v e n by v e r y s m a l l d i f f e r e n c e s o f t h e a u d i t o r y at-
Correspondence to: Dr. C. Pantev, University of Mfinster, Institute of Experimental Audiology, Clinical Research Unit Biomagnetism, Kardinal-von-Galen-Ring 111, D-4400 Miinster (Germany). Tel.: +49-251 836885; Fax: +49-251 836883. * This work was supported by grants from the Deutsche Forschungsgemeinschaft (Klinische Forschergruppe "Biomagnetismus und Biosignalanalyse") and from the Alexander yon Humboldt Stiftung. ** Alexander yon Humboldt research fellow at the Institute of Experimental Audiology, University of Miinster. Permanent address: Institute of Psychology of the Hungarian Academy of Sciences, Department of Psychophysiology, P.O. Box 398, H-1394 Budapest, Hungary.
t r i b u t e s b o t h d u r i n g an i g n o r e a n d a t t e n d c o n d i t i o n , t h e N 2 b is usually e v o k e d in an a t t e n d c o n d i t i o n only. S i n c e t h e first r e p o r t s on the M M N , s i m i l a r b r a i n r e s p o n s e s h a v e b e e n r e c o r d e d in passive o d d b a l l p a r a d i g m s d e s i g n e d to study c h a n g e s in f r e q u e n c y , intensity, d u r a t i o n , rise t i m e , i n t e r s t i m u l u s i n t e r v a l (ISI), p h o n e t i c s t r u c t u r e , a n d spatial l o c a t i o n o f the a c o u s t i c s o u r c e ( f o r a r e v i e w , see N~i~itfinen 19901. With the advent of multichannel recording of electric d a t a , e x p e r i m e n t s h a v e s t a r t e d to f o c u s o n the l o c a t i o n of the MMN generators. To localize the sources of the M M N , d i f f e r e n t analysis t e c h n i q u e s h a v e b e e n used. T h a t t h e M M N is c o m p o s e d o f two s e q u e n t i a l p h a s e s was c o n f i r m e d by a p p l y i n g b r a i n e l e c t r i c s o u r c e analysis ( S c h e r g et al. 19891. T h e early p h a s e o f t h e differe n c e w a v e f o r m was e l i c i t e d o n l y by a l a r g e d e v i a n c e , w h i c h t h e a u t h o r s i n t e r p r e t e d as b e i n g d u e to an e n h a n c e m e n t o f t h e s u p r a t e m p o r a l N1 g e n e r a t o r (see N~ifit~inen a n d P i c t o n 19871. T h e e x p e r i m e n t a l d a t a o f S c h e r g et al. (1989) a n d N o v a k et al. (19901 r e v e a l e d t h a t t h e s u p r a t e m p o r a l N I g e n e r a t o r is c o n s i d e r a b l y e n h a n c e d for t h e l a r g e f r e q u e n c y d e v i a t i o n w h i l e t h e small f r e q u e n c y c h a n g e h a d no m e a s u r a b l e i n f l u e n c e on it. O n l y t h e late p h a s e o f the d i f f e r e n c e w a v e f o r m , e l i c i t e d by b o t h a s m a l l e r a n d a l a r g e r d e v i a n c e , was i n t e r p r e t e d as a g e n u i n e M M N c o m p o n e n t .
MMF TO MINOR PITCH CHANGES
Two different MMN generators were reported by Giard et al. (1990). The applied current density analysis revealed a frontal generator with right hemisphere dominance and a larger contralateral supratemporal generator. The topographic mapping study of Novak et al. (1990) also distinguished early and late supratemporal subcomponents corresponding to Scherg's analysis (1989). A very detailed analysis of the scalp distribution of the MMN (Paavilainen et al. 1991) revealed two somewhat different MMN subcomponents. The early MMN was interpreted as a complex wave composed of frontal and temporal subcomponents, while the late phase of the MMN was interpreted as being generated in the secondary or association areas of the auditory cortex. First evoked magnetic field recordings (Hari et al. 1984; Sams et al. 1985) confirmed that the magnetic counterpart of the MMN (called MMNm or MMF) is generated in the auditory cortex on the supratemporal plane. Recent magnetic studies have shown that the MMF can be elicited by deviations in intensity (Lounasmaa et al. 1989) and in duration (Kaukoranta et al. 1989), and even by changes in a steady-state rhythm of stimuli with the same stimulus attributes (Hari 1990). The equivalent source of the MMF was found to be slightly different from the source of the magnetic counterpart of N1 (N100m). Recently Sams et al. (1991) reported similar locations of equivalent current dipoles (ECD) for 3 types of MMF: frequency, intensity, and duration. The source of the MMF elicited by frequency and intensity deviance was significantly anterior to that of the slow evoked field component N100m, elicited by the standards and by the deviants. Although it was not discussed, the reported difference in anterior-posterior direction of the ECD locations for N100m to the standards and to the frequency deviants was larger than the reported significant difference between the source location of the difference wave MMF and N100m (Sams et al. 1991). The organization of the generator sites of the MMF elicited by intensity and frequency deviations is less clear than that of the N100m sources corresponding to these auditory stimulus attributes (Pantev et al. 1988, 1989a,b). Interpretations of the early part of the deviant-standard difference waves in terms of an enhancement of tire supratemporal N1 generator (Scherg et al. 1989; Novak et al. 1990) have pointed to the significant influence of the degree of the stimulus deviance, especially in case of frequency deviation. The goals of this study were: (1) To separate the source of the MMF elicited by a very small frequency deviance from the source of the simultaneously generated N100m (termed M100). (2) To investigate the effect of stimulus order on the mismatch process to test predictions of the memory trace hypothesis that, if
539
all stimuli of the auditory input are compared with the existing memory trace, and if two different neural stimulus representations can exist in parallel, then the response to the standard following a deviant would also elicit a mismatch wave. The wave forms of the derived MMF were obtained by subtracting the wave forms of the responses to the standards from the wave form of the averaged responses to the deviants, a procedure which is typically used in ERP studies in order to extract a particular portion of a complex response. (3) To compare the source locations of the MMF in the derived waveform with those of the "identifiable" MMF components of the averaged response to the deviant (termed the native MMF).
Methods
(a) Subjects Nine subjects (4 females and 5 males) with no history of otological or neurological disorder and normal audiological status, ranging in age between 18 and 44, participated in this study. Informed consent was obtained from the subjects after the nature of the study was fully explained to them. One of the subjects (a female) was dominantly left-handed. The other eight were right-handed as determined with a modified handedness questionnaire of Annett (1967). Responses in the hemisphere contralateral to the dominant hand were investigated. The stimuli were delivered contralateral to the measured hemisphere, since previous studies have shown that the strongest auditory evoked magnetic fields (AEFs) are recorded over the hemisphere contralateral to the side of handedness (Elberling et al. 1981; Hoke 1988) and contralateral to the side of stimulation (Elberling et al. 1980, 1981; Reite et al. 1981; Pantev et al. 1986).
(b) Stimulation Two blocks of 1024 stimuli (test-retest) were presented to the subject. Each block consisted of two types of stimulus: standards (probability 90%) and deviants (probability 10%), delivered in random order with deviants followed by at least 3 standards, with a random ISI between 600 and 900 msec. The standard stimuli were 1000 Hz tonebursts with a duration of 50 msec and 5 msec rise and decay times (with cosine envelope), and an intensity of 60 dB n H L (normal hearing level). The deviant stimuli were tone bursts with the same envelope as that of the standards but with a deviant carrier frequency of 1050 Hz. The stimuli were presented to the subjects' ear through a non-magnetic and echo-free stimulus delivery system with an almost linear frequency characteristic (changes of _+4 dB in the range between 250 and 4000 Hz). During stimulus presentation the subjects were re-
540
V. CSEPE ET AL.
quested to analyse complex pictures of M.C. Escher, mounted on the wall in front of the subjects' field of view.
(c) Neuromagnetic measurements Magnetic data were recorded through a 37-channel biomagnetometer (Magnes TM, Biomagnetic Technolo-
C
gies) in the magnetically shielded room of the Institute of Experimental Audiology. In this instrument the detection coils are arranged in a circular array (diameter 144 mm) on the surface of a sphere (radius 122 mm), with their axes normal to the surface of the sphere. The distance between the centres of the coils is 22 ram, the coil diameter 20 mm. The spectral density of the
AEF: St all
AEF: St pre
MMF: D ev
AEF: St foll D
Fig. 1. Averaged wave forms (test-retest overplots), measured over the left hemisphere of one subject, in response to the standards preceding the ieviants (B), following the deviants (D), to all remaining standards of the block (A), and to the deviants (C). The most prominent deflection of he transient AEF, the M100 deflection, represents the ingoing (anterior) and the outgoing (posterior) evoked magnetic field. The second leflection in response to the deviants with the same polarity as wave M100 represents the MMF. The placement of the pick-up coils of the magnetic sensors relative to the head is shown in the insert in the middle of the figure.
MMF TO MINOR PITCH CHANGES intrinsic noise of each channel was between 5 and 7 fF/H~H7 (for more details see Pantev et al. 1991). The neuromagnetic field pattern was recorded over the supratemporal cortex. 1024 stimulus-related epochs of 500 msec with 100 msec prestimulus interval were recorded sequentially in two experimental runs (test/retest) with a passband of 0.1-200 Hz (sampling rate 860 Hz), then digitally filtered off-line between 0.1 and 30 Hz.
(d) ERP recordings In 5 of the 9 subjects ERPs were also recorded on two channels iFz, Cz) using gold-plated Grass electrodes. The earlobe contralateral to the side of stimulation served as reference, and the ground electrode was placed on the forehead. The same recording and filter settings were applied as in the magnetic recordings.
541 positions), orientation and amplitude of an equivalent current dipole (ECD) tangential to the surface of the model sphere (Geselowitz 1970; Sarvas 1987) were estimated for each point in time ( - 1 0 0 to 400 msec). The origin of the head-based coordinate system (determined by a sensor position indicator) was the midpoint between the preauricular points. The x-axis joined the origin to the nasion; the y-axis passed between the preauricular points with positive values towards the left preauricular point. The z-axis was perpendicular to the x-y plane. Correlations between the theoretical field generated by the model and the observed field were used to estimate the goodness of fit of the model parameters. Only estimates with a goodness of fit above 0.90 were analysed further. Statistical analysis of differences in the latencies and in the amplitudes of the M100 and MMF as well as of their ECD parameters in the original and derived fields (D, P, F, A, D-P, D-F, D-A) was carried out by applying 2-sided t tests for paired samples.
(e) On-line experimental control and off-line data analysis The ongoing data acquisition process was controlled by a separate VME-bus computer (Motorola 68030/ 68882), which received 8 arbitrarily selected magnetic channels as input. Its main tasks were to control the stimulus generator, to compute selective averages of the different sets of responses and of the derived wave forms, and to monitor the acquired and preprocessed data. Four selective averages were computed: the response to the deviants (D), responses to standards preceding (P) and following (F) the deviants, and the response to all remaining standards (A). Additionally, since the responses D, P, F and A were recorded in the same sessions and with exactly the same sensor array position, derived wave forms were computed by subtracting, from the D wave form, the P, F and A wave forms. For both the on-line experimental control and the off-line data analysis a fixed amplitude artifact rejection level was employed, discarding about 5 - 1 0 % of artifact-contaminated epochs because of eye blinks or movements. Since wave forms of the averaged responses recorded in the test and retest runs were reasonably similar for all subjects, individual grand averages of the two were used for further evaluation. Latencies and amplitudes of the M100 and MMF in the different average wave forms D, P, F and A and in the derived average wave forms D-P, D-F and D-A were analyzed for each subject for the 6 channels located nearest to the anterior and the posterior field maxima. A source analysis based on a single moving dipole model was applied to each of the obtained field distributions. For each subject a spherical model was fitted to the digitized head shape, and the location (x, y, z
Results
The wave forms obtained by selective averaging of the data (test-retest runs) are illustrated in Fig. 1. The wave forms of the transient A E F in response to the standards (P, F and A; Fig. 1A, B and D) are characterized by a prominent M100 deflection, preceded by an M50 and followed by an M200 of opposite polarity. Mean latencies and standard deviations for M100, computed over the 9 investigated subjects, were 105 _+ 15 msec at the anterior and 109 + 15 msec at the posterior extrema. Stimulus Onset is marked by a vertical line on the x-axis. As seen in Fig. 1, the S / N ratio of the A wave forms is distinctly better than that of the P, F and D responses. This is due to the fact that this average comprises approximately 7 times the number of epochs as in the P and F standards and the deviants. The responses to the deviants (D) are much more complex than the responses to the standards. The most pronounced deflection M100 (latency 99 _+ 10 msec anterior, 108_+ 11 msec posterior) is followed by the MMF (latency 205 + 34 msec anterior and 211 _+ 27 msec posterior). As mentioned above, the MMF was isolated from the M100 by subtracting the wave form of the responses to the standards (A, P, or F) from that of the response to the deviants (D). The application of the derived technique, however, requires a sufficiently high S / N ratio in both wave forms and a high similarity of the transient A E F deflections in the wave forms to be subtracted. These conditions can be achieved only if the fields related to both events are collected in a similar state of vigilance.
542
The S / N ratio of the wave forms being subtracted is not necessarily similar, because the state of vigilance may change during data acquisition. Since the P, D and F stimuli follow one after another, vigilance changes will have only a minor effect on the S / N ratio of the corresponding selective averages of the original (D, P and F) and the derived MMF wave forms D-P and D-F. On the other hand, for the derived MMF wave form D-A, the prerequisite of similarity is not fulfilled. The A wave form has a distinctly better S / N ratio, but is probably more influenced by long-term vigilance changes than the D, P and F wave forms. In order to examine the influence of both S / N ratio and changes in vigilance, the derived MMF wave forms D-A and D-P were compared for each subject. Though there is no significant amplitude difference between the D-A, D-P, D-F and Dev mismatch fields, significant differences ( P < 0.01) were found between the corresponding data in the anterior and posterior areas. The same trend was observed in the M100 amplitudes of the standards and of the deviants, but the differences did not reach the same level of significance. The latencies showed no significant differences at all. Very similar wave forms were obtained for the derived magnetic MMF and the derived electrical MMN (Fig. 2). The figure shows the derived wave forms obtained by D-A subtraction for one subject. The derived MMN is shown for two electrode positions (Cz and Fz), while the derived MMF displays the magnetic activity in several channels located along a line connecting the two field extrema. The magnetic and electric data obtained from all subjects showed similar wave forms except for one subject (M.G.), who was excluded from the data pool. Before addressing the question of the source location for the MMF, two basic questions have to be answered: first, do the ECD parameters of the M100 in the P, F, A and D wave forms differ significantly, and second, are there significant differences between the ECD parameters obtained for the 3 derived MMF wave forms D-A, D-P and D-F? These questions were studied for each subject and then compared statistically across subjects. Fig. 3 (top half) displays the field distribution patterns of M100 obtained from the A and D wave forms. The field patterns have the same structure and orientation, but a larger dipole moment (represented by the length of the arrows) for the deviants. In contrast, the patterns obtained from the native (D) and the derived MMF (D-A) do not exhibit large differences (Fig. 3, bottom half). It is, however, obvious that the source dipoles for the MMF are shifted slightly and rotated by about 10° counterclockwise compared to those of the M100. The second question, whether the source location of the derived MMF differs in the two cases D-P and
V. CSI~PE ET AL.
~ J~
Ch 10 Ch 3 Chl
~
Ch17ch 33 1O0 fT Cz Fz 4 #V
-0.1
0
0.1 0.2 0.3 Latency (sec)
0.4
0.5
Fig. 2. Comparison of the wave forms of the derived M M F and the MMN. The M M N represents the activity recorded from Cz and Fz whereas the M M F shows the magnetic activity of 5 channels located along a line connecting the two field extrema. The stimulus onset is marked by the arrow.
D-F, is important for both methodological and theoretical reasons. The distribution patterns for D-P and D-F responses from one subject are shown in Fig. 4, with that for the D-A case shown for comparison. Fig. 4A illustrates superimposed D-A, D-F and D-P wave forms from the test-retest sessions of one subject. It is rather difficult to distinguish the 3 individual wave forms. This observation is confirmed by the similar MMF distribution patterns for D-A, D-P and D-F (Fig. 4B, C and D). The analysis of the dipole parameters shown in Fig. 5 did not reveal significant differences in the source locations of the derived D-P and D-F MMFs in the latency range of the field maximum. Only a tendency for a higher dipole moment in the D-F case was observed. The most important questions to be answered by the present study are whether the derived MMF technique introduces unpredictable errors in location, and where the source of the derived MMF is located relative to that of wave M100. Averages of the individual x, y and z ECD coordinates during the 30-50 msec range around the field maxima of the M100, the native MMF and the derived MMF, respectively, were computed for each of the 8 subjects. These values were supposed to represent the
MMF TO MINOR, PITCH CHANGES
543
average "centre of activity" of the excited neuronal populations. The corresponding source locations for the native (filled circles) and for the derived MMF
(D-A) (empty circles) for all subjects are shown in Fig. 6A. The grand means across subjects are illustrated by large circles. Statistical evaluation of the data confirms
MIO0: St ,/,\\\4,(/ I.j . • ",
/`3(;
_,
.
,
.xX
\
-.
<,
~.
,'
t
M100: Dev
/`
A23
0
~
/`24
',',',L',%\\\\\\\\~
~!
.'" ."U i ': ,', ~.. '-, ' % \ ~ A32
"
j
"
,
,
,"
"
,,"
,
'
'
j
'
•
'
'
~,
\
$
\~
Z
"
,'
/
/
,/`~o~
/
!
~
1
x
MMF: D-A
MMF: Dev k Az
A/A ~1
2
•
,~"?o
A 1
½z
A1
AIII
"~
""
•
t
,/3o
,3~,, ;
t
",
;
,/ ,'
/`~
L ~.
Fig. 3. ]sofield contour maps corresponding to the peak of M ] 0 0 , computed for the responses to all remaining standards ( A ) and to the deviants
(B), and corresponding to the peak of MMF, computed for the responses to the deviants (C) and for the D-A wave form (D). The solid lines represent the outgoing, the dashed lines the ingoing field. The contour step is 20 fT. The arrows represent the location and the orientation of the estimated ECD, its length being proportional to the dipole moment amplitude.
544
v. cSEPE ET AL.
the visual impression that there is no significant difference between the source locations of the native and derived MMFs.
,'
Contrary to this, there is a clear-cut difference ( P < 0.01) for each of the 3 spat(a! ECD coordinates between the derived MMF and the M100. The source of
MMF: D ' A
MMF: D - A , D-P, D-F
--., e,-,,,',,.,,. : , ~ ~
,
". . . . . "
' I
~'
•
200 IT I
\ \ X . \ ~ ~
',, ',"', ' , " , ' X \ ~ "
, ; .,',,"..-X
,1
400 m s
X
MMF: D-P
')
t
~
~/
MMF: D-F D.
', ,:'~o
,,
I/
/
,
/
/
Fig. 4. Overplot of the grand averages of the D-A, D-F and D-P wave forms from the test-retest sessions of one subject (A), The isofidd contour maps computed for the M M F pea]< of the data shown in A are illustrated in B (derived field D-A), in C (derived field D-P), and in D (derived field D-F). The plot parameters are the same as in Fig. 3.
MMF TO M I N O R PITCtt C H A N G E S
,,.-.,,
545
40-
E < C v
¢.-
• .,~x
30-
AA~^•
z~ Q.DF ~zx~zx~
E o
•
20. o
Z
Q. .
QIDP
m
D
10 0.15
,
0.16
,
,
,
0.18
0.17
r
i
i
i
i
i
E
i
0.2
0.19
0.21
_
E 0
v
o x.DF
6-
Or')
[] y.DF
o
4-
o0 O C~
2-
° ~
N
x"
J--
--.,,.u
n
v z.DF
•
0.: -2: 0.i5
A - -
l
~
'
•
x.DP
•
y.DP
•
z.DP
w
,
0.16
i
~
i
i
0.17
~
,
,
,
j
j
0.18
,
J
0.19
J
J
J
,
0.2
i
J
i
0.21
Latency (sec) Fig. 5. Comparison between the source parameters of the derived D-P and D-F MMFs of one subject. The x, y and z positions and the dipole moment (Q) are shown as a function of time (160-200 msec after the stimulus onset).
the M M F is located between 1 and 15 m m more anterior (mean value across subjects: 7 mm), between 3 and 14 m m (mean: 7.5 mm) more medial and between 3 and 11 mm (mean: 6 m m ) more inferior than the source of the M100.
Discussion
These results demonstrate that a distinct MMF, corresponding in latency to the simultaneously recorded fronto-central MMN, can be elicited by an occasional minor pitch deviation. Source location analysis using a single moving dipole model showed the same spatial localization for the native M M F and for the derived MMFs (D-A, D-P, D-F). This finding demonstrates that the application of the single E C D model to the derived fields is appropriate when data are of sufficient quality due to coherent data acquisition and identical m e a s u r e m e n t positions. These consistent findings suggest that the neuromagnetic M M F can be well modelled as a distinct localized source. There are several studies of the electrical N100 demonstrating multiple generator sources (cf., N~i~it~inen and Picton 1987), which also may contribute to the magnetic M100. However, the dissociation of N100 and
M100 amplitude and latency at ISis longer than 4 sec (Hari et al. 1987) suggests that mainly supratemporal sources contribute to the generation of M100. Recent studies of Pantev et al. (1991), overlying A E F source locations onto M R I images, revealed that the M100 source can be localized with high reliability in the region of Heschl's gyrus and the transverse portion of the superior temporal gyrus. Since the cortical origin of wave M100 has been verified in extensive studies, it is justified to use it as an anatomical landmark for the source localization of the MMF. The sources of M100 showed no significant difference in their spatial location in response to standards and to deviants. The larger field amplitude in response to the deviants could be due to the excitation of non-refractory neurones of the same source. The derived deviance-related field can be regarded as the genuine M M F in agreement with results of brain electric source analysis of Scherg et al. (1989) and with that of mapping Study of Novak et al. (1990). On the other hand, the M M F observed in our study did not show any separation as reported by Paavilainen et al. (1991), although they also used a small pitch deviance, so that the early part of the M M N of 150 msec latency could not be due to a contamination from the enhancement of the supratemporal N100. The latency of the only
546
V. CSt~PE ET AL.
A
7
7
6
6
5
4
4
3
3 5
1 ant.post (crn)
~%lat (crn)
B
7
7
6
6
-g 5
4
4
3
3 5 6
me~lat(cnl)
v
1 ant.p0st (crn)
Fig. 6. A: estimated source locations for the native M M F in the responses to the deviants (filled circles) and for the derived M M F D-A (empty circles), obtained from all subjects. The means across subjects are illustrated by large symbols. B: source locations for the M M F D-A (empty circles) and for the MI00 of the standards (filled triangles). The means across all subjects are illustrated by large symbols, x-, y- and z-axis of the 3-dimensional plots represent the axes of the head-based coordinate system as described in the Methods section.
identifiable MMF peak in our study varied across subjects between 160 and 260 msec, which is in the latency range of the two subcomponents distinguished in Paavilainen et al.'s study (1991). The second negative phase of this MMN, having a scalp distribution different from that of the early phase, was interpreted as being due to
an additional, radial MMN subcomponent, generated in the secondary or association cortex. This late phase of the MMN, recordable with extremely lateral and posterior electrodes, supports the interpretation, obtained from animal data (Cs6pe et al. 1987), of generators of the frequency mismatch in the association cortex. In neuromagnetic measurements, however, the contribution of a radially oriented source to the MMF cannot be measured. The fact that the late phase of the MMN appears as second peak in the electric measurements only is incompatible with a primary auditory cortical generator for the late response. While the latency of the MMF did not differ in the posterior and anterior areas, the mean posterior field was significantly stronger than the anterior field. The same trend could be observed in the M100s elicited by the standards and deviants. We do not know whether this difference is real. However, measurements of the magnetic activity of a single artificial dipole implanted in a realistic head-shaped phantom at a known position showed a similar, non-symmetrical field distribution. Therefore, it is rather probable that the observed amplitude difference is due to the different orientation of the magnetic sensors with respect to the MMF source at the anterior and posterior part of the recording surface. The consistently similar locations of the ECD for the native and the 3 derived MMFs demonstrate that the different S / N ratio of the standards A, D and F does not significantly influence the localization results. The reasonable similarity of the MMF recorded in the test-retest runs supports the experimental findings of N~i~it~inen and Gaillard (1983) and N~i~it~inen et al. (1987) that the attenuation of the MMN amplitude is negligible during the course of the experiments. However, a trend of a higher dipole moment for the D-F difference wave form compared to the D-A or D-P difference wave forms could be observed. The memory trace hypothesis, proposing an MMN also being elicited by the standards following the deviants, would account for a smaller D-F derived field. Our findings demonstrate just the opposite, although the amplitude of the D-F MMF shows larger intersubject variability than those of the D-A and D-P MMFs. The higher dipole moment of the D-F MMF can be interpreted as being due to smaller field values to the F standards compared to that of the A and P standards contributing to a depart value of the dipole moment of the D-A and D-P MMFs. Our finding seems to contradict the previous results of Sams et al. (1984), showing that the standard stimulus immediately following the deviant elicits a small MMN, suggesting that the deviant had also produced its own trace. In our experimental paradigm at least 3 standards had to occur between two deviants and the magnitude of the deviant-standard pitch difference was one-fifth of that used by
MMF TO MINOR PITCH CHANGES
Sams et et al. (1984), which might explain the difference in the results between the two studies. The present data also suggest that even a very small stimulus deviance causes distinct effects in magnetic measurements. In the frequency MMF data of Sams et al. (1991), who used a larger frequency deviance (standard 1000 Hz, deviant 1500 Hz), a large difference in the M100 source location in the anterior-posterior direction can be recognized between responses to the standards and to the deviants. The MMF (with a reported peak latency of 128 msec) merged with the M100, so that the question arose whether or not the observed changes represent separate mismatch sources. It is possible that different sources might be activated by increasing the degree of deviance, making a reliable component separation difficult even in the magnetic measurements. The source localization based upon the derived wave forms seems to be more significant when M100 and MMF overlap in case of shortening of the MMF latency due to a large stimulus deviance. The ECD location of the MMF, compared to that of M100, demonstrated very consistent and significant differences for all 3 spatial location parameters. The location of the MMF, being anterior, medial and inferior relative to MI00, obviously represents a steady source in the auditory cortex. The clearly separated locations of the MMF and M100 sources suggest different origins in the human auditory cortex. The cytoarchitectonic study of Galaburda and Sanides (1980) revealed a medial and a lateral subdivision of the "granular cove" (p. 604) (superior temporal plane and the superior temporal gyrus). The medial subdivision, being very rich in callosal connections, might correspond to an MMF source. However, the topographic segregation of these subdivisions has not been confirmed by physiological methods. Without accompanying anatomical imaging, the estimated ECDs, representing the "centre of activity" of the neuronal population, give only information about the relative locations of the sources rather than about the contributing cytoarchitectonic areas. The obtained results can be explained by two functionally different and spatially separable neuronal populations in the supratemporal auditory cortex which might be active in early sensory analysis of simple tones and detection of minor pitch deviations. One of the two sources represents the supratemporal M100 generator, whereas the olher represents the neuronal population generating the MMF, which can be regarded as an evoked field correlated with an automatic "change detection" process. The spatial and temporal separation of the two sources seem to support and verify N~i~.t~inen's proposal on the basic processing of auditory stimuli (1990), especially one of his two suggested processing modes: "task-independent" sensory analysis which is composed of a transient-detector and a "per-
547
manent" feature-detector system. The transient-detector system is organized in a stimulus-specific manner, structured according to tonotopic, amplitopic (Pantev et al. 1988, 1989a,b) or other feature maps, and activates the Nl-generator process. The result of this sensory analysis is stored in the short-term acoustic memory and reinforced by repetitions of the same stimulus feature to which the different stimuli are compared. Our data clearly demonstrate that, whereas slightly deviant frequencies do not activate different M100 generators, small deviations are already sufficient to trigger a spatially distinct mismatch generator. The authors thank Drs. R. N~i~it~inen and S. Makeig for thoroughly reviewing an earlier version of this manuscript.
References Annett, M. The binomial distribution of right, mixed and left-handed. Quart. J. Exp. Psychol., 1967, 19: 327-333. Cs6pe, V., Karmos, G. and Molnfir, M. Evoked potential correlates of stimulus deviance during wakefulness and sleep in cat animal model of mismatch negativity. Electroenceph. clin. Neurophysiol., 1987, 66: 571-578. Elberling, C., Bak, C., Kofoed, B., Lebech, J. and Saermark, K. Auditory magnetic fields from the human brain. Scand. Audiol., 1980, 9: 185-190. Elberling, C., Bak, C., Kofoed, B., Lebech, J. and Saermark, K. Auditory magnetic field. Auditory magnetic field from the human cortex: influence of stimulus intensity. Scand. Audiol., 1981, 10: 203 -207. Ford, J.M., Roth, W.T. and Kopell, B.S. Auditory evoked potentials to unpredictable shifts in pitch. Psychophysiology, 1976a, 13: 32-39. Ford, J.M., Roth, W.T. and Kopell, B.S. Attention effects on auditory evoked potentials to infrequent events. Biol. Psychol., 1976b, 4: 65-77. Galaburda, A. and Sanides, F. Cytoarchitectonic organization of the human auditory cortex. J. Comp. Neurol., 1980, 190: 597-61(/. Geselowitz, D.B. On the magnetic field generated outside an inhomogeneous volume conductor by internal current sources. IEEE Trans. Magn., 1970, 6: 346-347. Giard, M.H., Perrin, P. and Pernier, J. Brain generators implicated in processing of auditory stimulus deviance: a topographic ERP study. Psychophysiology, 1990, 27: 627-640. Hari, R. The neuromagnetic method in the study of the human auditory cortex. In: F. Grandori, M. Hoke and G.L. Romani (Eds.), Auditory Evoked Magnetic Fields and Electric Potentials. Adv. Audiol., Vol. 6. Karger, Basel, 1990: 222-282. Hari, R., H~im~il~iinen, M., Ilmoniemi, R., Kaukoranta, E., Reinikainen, K., Salminen, J., Alho, K., N~i~it~inen, R. and Sams, M. Responses of the primary auditory cortex to pitch changes in a sequence of tone pips: neuromagnetic recordings in man. Neurosci. Len., 1984, 50: 127-132. Hari, R., Pelizzone, M., M~ikel~i, 1.P., H~illstr6m, J., Leinonen, L. and Lounasmaa, O.V. Neuromagnetic responses of the human auditory cortex to on- and off-sets of noise bursts. Audiology, 1987, 26: 31-43. Hoke, M. Auditory evoked magnetic fields. In: E. Ba~ar (Ed.), Dynamics of Cognitive and Sensory Processing in the Brain. Springer, Berlin, 1988:311-318. Kaukoranta, E., Sams, M., Hark R., H~im~il~iinen, M. and N~i~it~inen,
548 R. Reactions of human auditory cortex to changes in tone duration. Hearing Res., 1989, 41: 15-22. Lounasmaa, O., Hari, R., Joutsiniemi, S.L. and HiimS_liiinen, M. MultiSQUID recordings of human cerebral magnetic fields may give information about memory processes. Europhys. Lett., 1989, 9: 6113-6//8. N~iS.tiinen, R. The role of attention in auditory information processing as revealed by event-related potentials and other brain measures of cognitive function, Behav. Brain Sci., 1990, 13: 201-288. N~iiitS.nen, R. and Galliard, A.W.K. The N2 deflection of ERP and the orienting reflex. In: A.W.K. Gaillard and W. Ritter (Eds.), EEG Correlates of Information Processing: Theoretical Issues. Elsevier, Amsterdam, 1983: 119-141. N~iS_t~inen, R. and Picton, T.W. The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure. Psychophysiology, 1987, 24: 375-425. N~i~it~inen, R., Galliard, A.W.K. and M~intysalo, S. Early selective attention effect on evoked potential reinterpreted. Acta Psychol. (Amst.), 1978, 42: 313-329. Niiiit~inen, R., Paavilainen, P., Alho, K., Reinikainen, K. and Sams, M. The mismatch negativity to intensity changes in an auditory stimulus sequence, ln: R. Johnson, Jr., J.W, Rohrbaugh and R. Pa,asuraman (Eds.), Current Trends in Event-Related Potential Research. Electroenceph. clin. Neurophysiol., Suppl. 40. Elsevier, Amsterdam, 1987: 125-131. Novak, G.P., Ritter, W., Vaughan, Jr., H.G. and Wiznitzer, M.L. Differentiation of negative event-related potentials in an auditory discrimination task. Electroenceph. clin. Neurophysiol., 1990, 75: 255-275. Paavilainen, P, Alho, K., Reinikainen, K., Sams, M. and Nii~itS_nen, R. Right hemisphere dominance of different mismatch negativities. Electroenceph. clin. Neurophysiol., 1991, 78: 466-479. Pantev, C., Liitkenh6ner, B., Hoke, M. and Lehnertz, K. Comparison between simultaneously recorded auditory-evoked magnetic fields and potentials elicited by ipsilateral, contralateral and binaural tone-burst stimulation. Audiology, 1986, 25: 54-61. Pantev, C., Hoke, M, Lehnertz, K., Liitkenh6ner, B., Anogianakis, G. and Wittkowski, W. Tonotopic organization of the human auditory cortex revealed by transient auditory evoked magnetic fields. Electroenceph. clin. Neurophysiol., 1988, 69: 160-170.
V. CSI~PE ET AL. Pantev, C., Hoke, M., Liitkenh6ner, B. and Lehnertz, K. Tonotopic organization of the auditory cortex: pitch versus frequency representation. Science, 1989a, 246: 486-488. Pantev, C., Hoke, M., Lehnertz, K. and Liitkenh6ner, B. Neuromagnetic evidence of an amplitopic organization of the human cortex. Electroenceph. clin. Neurophysiol., 1989b, 72: 225-231. Pantev, C., Gallen, C., ftampson, S,, Buchanan, S. and Sobel, D. Reproducibility and validity of neuromagnetic source localization using a large array biomagnetometer. Am. J. EEG Technol., 1991, 31: 83-11/1. Reite, M., Zimmerman, J.T. and Zimmerman, J.E. Magnetic auditory evoked fields: interhemispheric asymmetry. Electroenceph. clin. Neurophysiol., 1981, 5[: 388-392. Sams, M., Alho, K. and N~iS_t~inen,R. Sequential effects on the ERP in discriminating two stimuli. Biol. Psychol., 1983, 17: 41-58. Sams, M., Alho, K. and Nfi'atfinen, R. Short-term habituation and dishabituation of the mismatch negativity of the ERP. Psychophysiology, 1984, 21: 434-441. Sams, M., H~imS_l/iinen, M., Antervo, A., Kaukoranta, E.. Reinikainen, K. and Hari, R. Cerebral neuromagnetic responses ew)ked by short auditory stimuli. Electroenceph. clin. Neurophysiol., 1985, 61: 254-266. Sams, M., Kaukoranta, E., H~imiiliiinen, M. and NS_iit~inen, R. Cortical activity elicited by changes in auditory stimuli: different sources for the magnetic N1011m and mismatch responses. Psychophysiology, 1991, 28: 21-29. Sarvas, J. Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Phys. Med. Biol., 1987, 32: 11-22. Scherg, M., Vajsar, J. and Picton, T.W. A source analysis of the human auditory evoked potentials. J. Cogn. Neurosci., 1989, 1: 336-355. Simson, R., Vaughan, Jr., H.G. and Ritter, W. The scalp topography of potentials in auditory and visual discrimination tasks. Elcctroenceph, clin. Neurophysiol., 1977, 42: 528-535. Squires, K.C., Donchin, E., Herning, R.I. and McCarthy, G. On the influence of task relevance and stimulus probability on event-related stimulus components. Electroenceph. clin. Neurophysiol., 1977, 42: 1-14.
Electroenccphalography and clinical Neurophysiology, 84 (1992) 538-548 ~,~'~1992 Elsevier Scientific Publishers Ireland, Ltd. [1168-5597/92/$05.011
EVOPOT 925115
Evoked magnetic responses of the human auditory cortex to minor pitch changes: localization of the m i s m a t c h field * V. Cs6pe **, C. Pantev, M. Hoke, S. Hampson and B. Ross Institute ~ff"Experimet~tal Audiology, Unit'ersity of" Miinster, Miinster (Germany) (Accepted for publication: 23 July 1992)
Summary The neuromagnetic source Iocalizations of the auditory M 100 and the mismatch field (MMF) were studied using a large-array biomagnetometer. Standard tones of 1000 Hz and deviant tones of 11150 Hz were delivered with 90% and 10'~ probability, respectively. Wave forms of the derived MMF were computed by examining difference wave forms between the responses to the deviants and the responses to the standards preceding (D-P) and following (D-F) the deviants as well as to all remaining standards (D-A). The subset of standards preceding the deviants was used for a more realistic comparison with the set of deviants (having the same number of epochs and a similar signal-to-noise ratio), while the subset of standards following the deviants served to answer the question whether those standards also elicit an MMF. The MMF deflections were compared with each other, with the "'native" MMF occurring in response to the deviants, and with wave M 100. (The MMF as it appears in the unprocessed response to the deviants was termed "'native" for an easy distinction from the "derived" MMF.) Our results demonstrate a distinct MMF deflection, corresponding in latency to the simultaneously recorded fronto-central electrical MMN. Source analysis, using a single moving dipole model, showed the same spatial localization for the native MMF and for the different derived MMFs. The MMF source location turned out to be significantly anterior, medial and inferior relative to the sources of the M 1110. The present data also demonstrate that a minor frequency deviation may not activate measurably different MI00 generators, yet be sufficient It) trigger the nearby but spatially distinct mismatch generator. Key words: Auditory cortex; Auditory evoked magnetic field; Mismatch field; Derived mismatch field; Neuromagnetic source localization
A negalive component of the event-related brain potential (ERP), called the mismatch negativity ( M M N ) , d e s c r i b e d first by N~i~it~inen et al. (1978), o c c u r s in r e s p o n s e to a d e v i a n t s t i m u l u s p r e s e n t e d , i n an o t h e r w i s e h o m o g e n e o u s s e q u e n c e o f s t a n d a r d stimuli. T h e s e a u t h o r s i s o l a t e d t h e M M N f r o m w a v e N 2 d e s c r i b e d in m a n y p r e v i o u s s t u d i e s ( F o r d et al. 1976a,b; S i m s o n et al. 1977; S q u i r e s et al. 1977). L a t e r , N~i~it~inen a n d his c o - w o r k e r s (N~i~it~inen a n d G a i l l a r d 1983) sugg e s t e d t h a t w a v e N 2 is c o m p o s e d o f two c o m p o n e n t s , t h e N 2 b a n d M M N . W h i l e t h e M M N c a n be e l i c i t e d e v e n by v e r y s m a l l d i f f e r e n c e s o f t h e a u d i t o r y at-
Correspondence to: Dr. C. Pantev, University of Mfinster, Institute of Experimental Audiology, Clinical Research Unit Biomagnetism, Kardinal-von-Galen-Ring 111, D-4400 Miinster (Germany). Tel.: +49-251 836885; Fax: +49-251 836883. * This work was supported by grants from the Deutsche Forschungsgemeinschaft (Klinische Forschergruppe "Biomagnetismus und Biosignalanalyse") and from the Alexander yon Humboldt Stiftung. ** Alexander yon Humboldt research fellow at the Institute of Experimental Audiology, University of Miinster. Permanent address: Institute of Psychology of the Hungarian Academy of Sciences, Department of Psychophysiology, P.O. Box 398, H-1394 Budapest, Hungary.
t r i b u t e s b o t h d u r i n g an i g n o r e a n d a t t e n d c o n d i t i o n , t h e N 2 b is usually e v o k e d in an a t t e n d c o n d i t i o n only. S i n c e t h e first r e p o r t s on the M M N , s i m i l a r b r a i n r e s p o n s e s h a v e b e e n r e c o r d e d in passive o d d b a l l p a r a d i g m s d e s i g n e d to study c h a n g e s in f r e q u e n c y , intensity, d u r a t i o n , rise t i m e , i n t e r s t i m u l u s i n t e r v a l (ISI), p h o n e t i c s t r u c t u r e , a n d spatial l o c a t i o n o f the a c o u s t i c s o u r c e ( f o r a r e v i e w , see N~i~itfinen 19901. With the advent of multichannel recording of electric d a t a , e x p e r i m e n t s h a v e s t a r t e d to f o c u s o n the l o c a t i o n of the MMN generators. To localize the sources of the M M N , d i f f e r e n t analysis t e c h n i q u e s h a v e b e e n used. T h a t t h e M M N is c o m p o s e d o f two s e q u e n t i a l p h a s e s was c o n f i r m e d by a p p l y i n g b r a i n e l e c t r i c s o u r c e analysis ( S c h e r g et al. 19891. T h e early p h a s e o f t h e differe n c e w a v e f o r m was e l i c i t e d o n l y by a l a r g e d e v i a n c e , w h i c h t h e a u t h o r s i n t e r p r e t e d as b e i n g d u e to an e n h a n c e m e n t o f t h e s u p r a t e m p o r a l N1 g e n e r a t o r (see N~ifit~inen a n d P i c t o n 19871. T h e e x p e r i m e n t a l d a t a o f S c h e r g et al. (1989) a n d N o v a k et al. (19901 r e v e a l e d t h a t t h e s u p r a t e m p o r a l N I g e n e r a t o r is c o n s i d e r a b l y e n h a n c e d for t h e l a r g e f r e q u e n c y d e v i a t i o n w h i l e t h e small f r e q u e n c y c h a n g e h a d no m e a s u r a b l e i n f l u e n c e on it. O n l y t h e late p h a s e o f the d i f f e r e n c e w a v e f o r m , e l i c i t e d by b o t h a s m a l l e r a n d a l a r g e r d e v i a n c e , was i n t e r p r e t e d as a g e n u i n e M M N c o m p o n e n t .
MMF TO MINOR PITCH CHANGES
Two different MMN generators were reported by Giard et al. (1990). The applied current density analysis revealed a frontal generator with right hemisphere dominance and a larger contralateral supratemporal generator. The topographic mapping study of Novak et al. (1990) also distinguished early and late supratemporal subcomponents corresponding to Scherg's analysis (1989). A very detailed analysis of the scalp distribution of the MMN (Paavilainen et al. 1991) revealed two somewhat different MMN subcomponents. The early MMN was interpreted as a complex wave composed of frontal and temporal subcomponents, while the late phase of the MMN was interpreted as being generated in the secondary or association areas of the auditory cortex. First evoked magnetic field recordings (Hari et al. 1984; Sams et al. 1985) confirmed that the magnetic counterpart of the MMN (called MMNm or MMF) is generated in the auditory cortex on the supratemporal plane. Recent magnetic studies have shown that the MMF can be elicited by deviations in intensity (Lounasmaa et al. 1989) and in duration (Kaukoranta et al. 1989), and even by changes in a steady-state rhythm of stimuli with the same stimulus attributes (Hari 1990). The equivalent source of the MMF was found to be slightly different from the source of the magnetic counterpart of N1 (N100m). Recently Sams et al. (1991) reported similar locations of equivalent current dipoles (ECD) for 3 types of MMF: frequency, intensity, and duration. The source of the MMF elicited by frequency and intensity deviance was significantly anterior to that of the slow evoked field component N100m, elicited by the standards and by the deviants. Although it was not discussed, the reported difference in anterior-posterior direction of the ECD locations for N100m to the standards and to the frequency deviants was larger than the reported significant difference between the source location of the difference wave MMF and N100m (Sams et al. 1991). The organization of the generator sites of the MMF elicited by intensity and frequency deviations is less clear than that of the N100m sources corresponding to these auditory stimulus attributes (Pantev et al. 1988, 1989a,b). Interpretations of the early part of the deviant-standard difference waves in terms of an enhancement of tire supratemporal N1 generator (Scherg et al. 1989; Novak et al. 1990) have pointed to the significant influence of the degree of the stimulus deviance, especially in case of frequency deviation. The goals of this study were: (1) To separate the source of the MMF elicited by a very small frequency deviance from the source of the simultaneously generated N100m (termed M100). (2) To investigate the effect of stimulus order on the mismatch process to test predictions of the memory trace hypothesis that, if
539
all stimuli of the auditory input are compared with the existing memory trace, and if two different neural stimulus representations can exist in parallel, then the response to the standard following a deviant would also elicit a mismatch wave. The wave forms of the derived MMF were obtained by subtracting the wave forms of the responses to the standards from the wave form of the averaged responses to the deviants, a procedure which is typically used in ERP studies in order to extract a particular portion of a complex response. (3) To compare the source locations of the MMF in the derived waveform with those of the "identifiable" MMF components of the averaged response to the deviant (termed the native MMF).
Methods
(a) Subjects Nine subjects (4 females and 5 males) with no history of otological or neurological disorder and normal audiological status, ranging in age between 18 and 44, participated in this study. Informed consent was obtained from the subjects after the nature of the study was fully explained to them. One of the subjects (a female) was dominantly left-handed. The other eight were right-handed as determined with a modified handedness questionnaire of Annett (1967). Responses in the hemisphere contralateral to the dominant hand were investigated. The stimuli were delivered contralateral to the measured hemisphere, since previous studies have shown that the strongest auditory evoked magnetic fields (AEFs) are recorded over the hemisphere contralateral to the side of handedness (Elberling et al. 1981; Hoke 1988) and contralateral to the side of stimulation (Elberling et al. 1980, 1981; Reite et al. 1981; Pantev et al. 1986).
(b) Stimulation Two blocks of 1024 stimuli (test-retest) were presented to the subject. Each block consisted of two types of stimulus: standards (probability 90%) and deviants (probability 10%), delivered in random order with deviants followed by at least 3 standards, with a random ISI between 600 and 900 msec. The standard stimuli were 1000 Hz tonebursts with a duration of 50 msec and 5 msec rise and decay times (with cosine envelope), and an intensity of 60 dB n H L (normal hearing level). The deviant stimuli were tone bursts with the same envelope as that of the standards but with a deviant carrier frequency of 1050 Hz. The stimuli were presented to the subjects' ear through a non-magnetic and echo-free stimulus delivery system with an almost linear frequency characteristic (changes of _+4 dB in the range between 250 and 4000 Hz). During stimulus presentation the subjects were re-
540
V. CSEPE ET AL.
quested to analyse complex pictures of M.C. Escher, mounted on the wall in front of the subjects' field of view.
(c) Neuromagnetic measurements Magnetic data were recorded through a 37-channel biomagnetometer (Magnes TM, Biomagnetic Technolo-
C
gies) in the magnetically shielded room of the Institute of Experimental Audiology. In this instrument the detection coils are arranged in a circular array (diameter 144 mm) on the surface of a sphere (radius 122 mm), with their axes normal to the surface of the sphere. The distance between the centres of the coils is 22 ram, the coil diameter 20 mm. The spectral density of the
AEF: St all
AEF: St pre
MMF: D ev
AEF: St foll D
Fig. 1. Averaged wave forms (test-retest overplots), measured over the left hemisphere of one subject, in response to the standards preceding the ieviants (B), following the deviants (D), to all remaining standards of the block (A), and to the deviants (C). The most prominent deflection of he transient AEF, the M100 deflection, represents the ingoing (anterior) and the outgoing (posterior) evoked magnetic field. The second leflection in response to the deviants with the same polarity as wave M100 represents the MMF. The placement of the pick-up coils of the magnetic sensors relative to the head is shown in the insert in the middle of the figure.
MMF TO MINOR PITCH CHANGES intrinsic noise of each channel was between 5 and 7 fF/H~H7 (for more details see Pantev et al. 1991). The neuromagnetic field pattern was recorded over the supratemporal cortex. 1024 stimulus-related epochs of 500 msec with 100 msec prestimulus interval were recorded sequentially in two experimental runs (test/retest) with a passband of 0.1-200 Hz (sampling rate 860 Hz), then digitally filtered off-line between 0.1 and 30 Hz.
(d) ERP recordings In 5 of the 9 subjects ERPs were also recorded on two channels iFz, Cz) using gold-plated Grass electrodes. The earlobe contralateral to the side of stimulation served as reference, and the ground electrode was placed on the forehead. The same recording and filter settings were applied as in the magnetic recordings.
541 positions), orientation and amplitude of an equivalent current dipole (ECD) tangential to the surface of the model sphere (Geselowitz 1970; Sarvas 1987) were estimated for each point in time ( - 1 0 0 to 400 msec). The origin of the head-based coordinate system (determined by a sensor position indicator) was the midpoint between the preauricular points. The x-axis joined the origin to the nasion; the y-axis passed between the preauricular points with positive values towards the left preauricular point. The z-axis was perpendicular to the x-y plane. Correlations between the theoretical field generated by the model and the observed field were used to estimate the goodness of fit of the model parameters. Only estimates with a goodness of fit above 0.90 were analysed further. Statistical analysis of differences in the latencies and in the amplitudes of the M100 and MMF as well as of their ECD parameters in the original and derived fields (D, P, F, A, D-P, D-F, D-A) was carried out by applying 2-sided t tests for paired samples.
(e) On-line experimental control and off-line data analysis The ongoing data acquisition process was controlled by a separate VME-bus computer (Motorola 68030/ 68882), which received 8 arbitrarily selected magnetic channels as input. Its main tasks were to control the stimulus generator, to compute selective averages of the different sets of responses and of the derived wave forms, and to monitor the acquired and preprocessed data. Four selective averages were computed: the response to the deviants (D), responses to standards preceding (P) and following (F) the deviants, and the response to all remaining standards (A). Additionally, since the responses D, P, F and A were recorded in the same sessions and with exactly the same sensor array position, derived wave forms were computed by subtracting, from the D wave form, the P, F and A wave forms. For both the on-line experimental control and the off-line data analysis a fixed amplitude artifact rejection level was employed, discarding about 5 - 1 0 % of artifact-contaminated epochs because of eye blinks or movements. Since wave forms of the averaged responses recorded in the test and retest runs were reasonably similar for all subjects, individual grand averages of the two were used for further evaluation. Latencies and amplitudes of the M100 and MMF in the different average wave forms D, P, F and A and in the derived average wave forms D-P, D-F and D-A were analyzed for each subject for the 6 channels located nearest to the anterior and the posterior field maxima. A source analysis based on a single moving dipole model was applied to each of the obtained field distributions. For each subject a spherical model was fitted to the digitized head shape, and the location (x, y, z
Results
The wave forms obtained by selective averaging of the data (test-retest runs) are illustrated in Fig. 1. The wave forms of the transient A E F in response to the standards (P, F and A; Fig. 1A, B and D) are characterized by a prominent M100 deflection, preceded by an M50 and followed by an M200 of opposite polarity. Mean latencies and standard deviations for M100, computed over the 9 investigated subjects, were 105 _+ 15 msec at the anterior and 109 + 15 msec at the posterior extrema. Stimulus Onset is marked by a vertical line on the x-axis. As seen in Fig. 1, the S / N ratio of the A wave forms is distinctly better than that of the P, F and D responses. This is due to the fact that this average comprises approximately 7 times the number of epochs as in the P and F standards and the deviants. The responses to the deviants (D) are much more complex than the responses to the standards. The most pronounced deflection M100 (latency 99 _+ 10 msec anterior, 108_+ 11 msec posterior) is followed by the MMF (latency 205 + 34 msec anterior and 211 _+ 27 msec posterior). As mentioned above, the MMF was isolated from the M100 by subtracting the wave form of the responses to the standards (A, P, or F) from that of the response to the deviants (D). The application of the derived technique, however, requires a sufficiently high S / N ratio in both wave forms and a high similarity of the transient A E F deflections in the wave forms to be subtracted. These conditions can be achieved only if the fields related to both events are collected in a similar state of vigilance.
542
The S / N ratio of the wave forms being subtracted is not necessarily similar, because the state of vigilance may change during data acquisition. Since the P, D and F stimuli follow one after another, vigilance changes will have only a minor effect on the S / N ratio of the corresponding selective averages of the original (D, P and F) and the derived MMF wave forms D-P and D-F. On the other hand, for the derived MMF wave form D-A, the prerequisite of similarity is not fulfilled. The A wave form has a distinctly better S / N ratio, but is probably more influenced by long-term vigilance changes than the D, P and F wave forms. In order to examine the influence of both S / N ratio and changes in vigilance, the derived MMF wave forms D-A and D-P were compared for each subject. Though there is no significant amplitude difference between the D-A, D-P, D-F and Dev mismatch fields, significant differences ( P < 0.01) were found between the corresponding data in the anterior and posterior areas. The same trend was observed in the M100 amplitudes of the standards and of the deviants, but the differences did not reach the same level of significance. The latencies showed no significant differences at all. Very similar wave forms were obtained for the derived magnetic MMF and the derived electrical MMN (Fig. 2). The figure shows the derived wave forms obtained by D-A subtraction for one subject. The derived MMN is shown for two electrode positions (Cz and Fz), while the derived MMF displays the magnetic activity in several channels located along a line connecting the two field extrema. The magnetic and electric data obtained from all subjects showed similar wave forms except for one subject (M.G.), who was excluded from the data pool. Before addressing the question of the source location for the MMF, two basic questions have to be answered: first, do the ECD parameters of the M100 in the P, F, A and D wave forms differ significantly, and second, are there significant differences between the ECD parameters obtained for the 3 derived MMF wave forms D-A, D-P and D-F? These questions were studied for each subject and then compared statistically across subjects. Fig. 3 (top half) displays the field distribution patterns of M100 obtained from the A and D wave forms. The field patterns have the same structure and orientation, but a larger dipole moment (represented by the length of the arrows) for the deviants. In contrast, the patterns obtained from the native (D) and the derived MMF (D-A) do not exhibit large differences (Fig. 3, bottom half). It is, however, obvious that the source dipoles for the MMF are shifted slightly and rotated by about 10° counterclockwise compared to those of the M100. The second question, whether the source location of the derived MMF differs in the two cases D-P and
V. CSI~PE ET AL.
~ J~
Ch 10 Ch 3 Chl
~
Ch17ch 33 1O0 fT Cz Fz 4 #V
-0.1
0
0.1 0.2 0.3 Latency (sec)
0.4
0.5
Fig. 2. Comparison of the wave forms of the derived M M F and the MMN. The M M N represents the activity recorded from Cz and Fz whereas the M M F shows the magnetic activity of 5 channels located along a line connecting the two field extrema. The stimulus onset is marked by the arrow.
D-F, is important for both methodological and theoretical reasons. The distribution patterns for D-P and D-F responses from one subject are shown in Fig. 4, with that for the D-A case shown for comparison. Fig. 4A illustrates superimposed D-A, D-F and D-P wave forms from the test-retest sessions of one subject. It is rather difficult to distinguish the 3 individual wave forms. This observation is confirmed by the similar MMF distribution patterns for D-A, D-P and D-F (Fig. 4B, C and D). The analysis of the dipole parameters shown in Fig. 5 did not reveal significant differences in the source locations of the derived D-P and D-F MMFs in the latency range of the field maximum. Only a tendency for a higher dipole moment in the D-F case was observed. The most important questions to be answered by the present study are whether the derived MMF technique introduces unpredictable errors in location, and where the source of the derived MMF is located relative to that of wave M100. Averages of the individual x, y and z ECD coordinates during the 30-50 msec range around the field maxima of the M100, the native MMF and the derived MMF, respectively, were computed for each of the 8 subjects. These values were supposed to represent the
MMF TO MINOR, PITCH CHANGES
543
average "centre of activity" of the excited neuronal populations. The corresponding source locations for the native (filled circles) and for the derived MMF
(D-A) (empty circles) for all subjects are shown in Fig. 6A. The grand means across subjects are illustrated by large circles. Statistical evaluation of the data confirms
MIO0: St ,/,\\\4,(/ I.j . • ",
/`3(;
_,
.
,
.xX
\
-.
<,
~.
,'
t
M100: Dev
/`
A23
0
~
/`24
',',',L',%\\\\\\\\~
~!
.'" ."U i ': ,', ~.. '-, ' % \ ~ A32
"
j
"
,
,
,"
"
,,"
,
'
'
j
'
•
'
'
~,
\
$
\~
Z
"
,'
/
/
,/`~o~
/
!
~
1
x
MMF: D-A
MMF: Dev k Az
A/A ~1
2
•
,~"?o
A 1
½z
A1
AIII
"~
""
•
t
,/3o
,3~,, ;
t
",
;
,/ ,'
/`~
L ~.
Fig. 3. ]sofield contour maps corresponding to the peak of M ] 0 0 , computed for the responses to all remaining standards ( A ) and to the deviants
(B), and corresponding to the peak of MMF, computed for the responses to the deviants (C) and for the D-A wave form (D). The solid lines represent the outgoing, the dashed lines the ingoing field. The contour step is 20 fT. The arrows represent the location and the orientation of the estimated ECD, its length being proportional to the dipole moment amplitude.
544
v. cSEPE ET AL.
the visual impression that there is no significant difference between the source locations of the native and derived MMFs.
,'
Contrary to this, there is a clear-cut difference ( P < 0.01) for each of the 3 spat(a! ECD coordinates between the derived MMF and the M100. The source of
MMF: D ' A
MMF: D - A , D-P, D-F
--., e,-,,,',,.,,. : , ~ ~
,
". . . . . "
' I
~'
•
200 IT I
\ \ X . \ ~ ~
',, ',"', ' , " , ' X \ ~ "
, ; .,',,"..-X
,1
400 m s
X
MMF: D-P
')
t
~
~/
MMF: D-F D.
', ,:'~o
,,
I/
/
,
/
/
Fig. 4. Overplot of the grand averages of the D-A, D-F and D-P wave forms from the test-retest sessions of one subject (A), The isofidd contour maps computed for the M M F pea]< of the data shown in A are illustrated in B (derived field D-A), in C (derived field D-P), and in D (derived field D-F). The plot parameters are the same as in Fig. 3.
MMF TO M I N O R PITCtt C H A N G E S
,,.-.,,
545
40-
E < C v
¢.-
• .,~x
30-
AA~^•
z~ Q.DF ~zx~zx~
E o
•
20. o
Z
Q. .
QIDP
m
D
10 0.15
,
0.16
,
,
,
0.18
0.17
r
i
i
i
i
i
E
i
0.2
0.19
0.21
_
E 0
v
o x.DF
6-
Or')
[] y.DF
o
4-
o0 O C~
2-
° ~
N
x"
J--
--.,,.u
n
v z.DF
•
0.: -2: 0.i5
A - -
l
~
'
•
x.DP
•
y.DP
•
z.DP
w
,
0.16
i
~
i
i
0.17
~
,
,
,
j
j
0.18
,
J
0.19
J
J
J
,
0.2
i
J
i
0.21
Latency (sec) Fig. 5. Comparison between the source parameters of the derived D-P and D-F MMFs of one subject. The x, y and z positions and the dipole moment (Q) are shown as a function of time (160-200 msec after the stimulus onset).
the M M F is located between 1 and 15 m m more anterior (mean value across subjects: 7 mm), between 3 and 14 m m (mean: 7.5 mm) more medial and between 3 and 11 mm (mean: 6 m m ) more inferior than the source of the M100.
Discussion
These results demonstrate that a distinct MMF, corresponding in latency to the simultaneously recorded fronto-central MMN, can be elicited by an occasional minor pitch deviation. Source location analysis using a single moving dipole model showed the same spatial localization for the native M M F and for the derived MMFs (D-A, D-P, D-F). This finding demonstrates that the application of the single E C D model to the derived fields is appropriate when data are of sufficient quality due to coherent data acquisition and identical m e a s u r e m e n t positions. These consistent findings suggest that the neuromagnetic M M F can be well modelled as a distinct localized source. There are several studies of the electrical N100 demonstrating multiple generator sources (cf., N~i~it~inen and Picton 1987), which also may contribute to the magnetic M100. However, the dissociation of N100 and
M100 amplitude and latency at ISis longer than 4 sec (Hari et al. 1987) suggests that mainly supratemporal sources contribute to the generation of M100. Recent studies of Pantev et al. (1991), overlying A E F source locations onto M R I images, revealed that the M100 source can be localized with high reliability in the region of Heschl's gyrus and the transverse portion of the superior temporal gyrus. Since the cortical origin of wave M100 has been verified in extensive studies, it is justified to use it as an anatomical landmark for the source localization of the MMF. The sources of M100 showed no significant difference in their spatial location in response to standards and to deviants. The larger field amplitude in response to the deviants could be due to the excitation of non-refractory neurones of the same source. The derived deviance-related field can be regarded as the genuine M M F in agreement with results of brain electric source analysis of Scherg et al. (1989) and with that of mapping Study of Novak et al. (1990). On the other hand, the M M F observed in our study did not show any separation as reported by Paavilainen et al. (1991), although they also used a small pitch deviance, so that the early part of the M M N of 150 msec latency could not be due to a contamination from the enhancement of the supratemporal N100. The latency of the only
546
V. CSt~PE ET AL.
A
7
7
6
6
5
4
4
3
3 5
1 ant.post (crn)
~%lat (crn)
B
7
7
6
6
-g 5
4
4
3
3 5 6
me~lat(cnl)
v
1 ant.p0st (crn)
Fig. 6. A: estimated source locations for the native M M F in the responses to the deviants (filled circles) and for the derived M M F D-A (empty circles), obtained from all subjects. The means across subjects are illustrated by large symbols. B: source locations for the M M F D-A (empty circles) and for the MI00 of the standards (filled triangles). The means across all subjects are illustrated by large symbols, x-, y- and z-axis of the 3-dimensional plots represent the axes of the head-based coordinate system as described in the Methods section.
identifiable MMF peak in our study varied across subjects between 160 and 260 msec, which is in the latency range of the two subcomponents distinguished in Paavilainen et al.'s study (1991). The second negative phase of this MMN, having a scalp distribution different from that of the early phase, was interpreted as being due to
an additional, radial MMN subcomponent, generated in the secondary or association cortex. This late phase of the MMN, recordable with extremely lateral and posterior electrodes, supports the interpretation, obtained from animal data (Cs6pe et al. 1987), of generators of the frequency mismatch in the association cortex. In neuromagnetic measurements, however, the contribution of a radially oriented source to the MMF cannot be measured. The fact that the late phase of the MMN appears as second peak in the electric measurements only is incompatible with a primary auditory cortical generator for the late response. While the latency of the MMF did not differ in the posterior and anterior areas, the mean posterior field was significantly stronger than the anterior field. The same trend could be observed in the M100s elicited by the standards and deviants. We do not know whether this difference is real. However, measurements of the magnetic activity of a single artificial dipole implanted in a realistic head-shaped phantom at a known position showed a similar, non-symmetrical field distribution. Therefore, it is rather probable that the observed amplitude difference is due to the different orientation of the magnetic sensors with respect to the MMF source at the anterior and posterior part of the recording surface. The consistently similar locations of the ECD for the native and the 3 derived MMFs demonstrate that the different S / N ratio of the standards A, D and F does not significantly influence the localization results. The reasonable similarity of the MMF recorded in the test-retest runs supports the experimental findings of N~i~it~inen and Gaillard (1983) and N~i~it~inen et al. (1987) that the attenuation of the MMN amplitude is negligible during the course of the experiments. However, a trend of a higher dipole moment for the D-F difference wave form compared to the D-A or D-P difference wave forms could be observed. The memory trace hypothesis, proposing an MMN also being elicited by the standards following the deviants, would account for a smaller D-F derived field. Our findings demonstrate just the opposite, although the amplitude of the D-F MMF shows larger intersubject variability than those of the D-A and D-P MMFs. The higher dipole moment of the D-F MMF can be interpreted as being due to smaller field values to the F standards compared to that of the A and P standards contributing to a depart value of the dipole moment of the D-A and D-P MMFs. Our finding seems to contradict the previous results of Sams et al. (1984), showing that the standard stimulus immediately following the deviant elicits a small MMN, suggesting that the deviant had also produced its own trace. In our experimental paradigm at least 3 standards had to occur between two deviants and the magnitude of the deviant-standard pitch difference was one-fifth of that used by
MMF TO MINOR PITCH CHANGES
Sams et et al. (1984), which might explain the difference in the results between the two studies. The present data also suggest that even a very small stimulus deviance causes distinct effects in magnetic measurements. In the frequency MMF data of Sams et al. (1991), who used a larger frequency deviance (standard 1000 Hz, deviant 1500 Hz), a large difference in the M100 source location in the anterior-posterior direction can be recognized between responses to the standards and to the deviants. The MMF (with a reported peak latency of 128 msec) merged with the M100, so that the question arose whether or not the observed changes represent separate mismatch sources. It is possible that different sources might be activated by increasing the degree of deviance, making a reliable component separation difficult even in the magnetic measurements. The source localization based upon the derived wave forms seems to be more significant when M100 and MMF overlap in case of shortening of the MMF latency due to a large stimulus deviance. The ECD location of the MMF, compared to that of M100, demonstrated very consistent and significant differences for all 3 spatial location parameters. The location of the MMF, being anterior, medial and inferior relative to MI00, obviously represents a steady source in the auditory cortex. The clearly separated locations of the MMF and M100 sources suggest different origins in the human auditory cortex. The cytoarchitectonic study of Galaburda and Sanides (1980) revealed a medial and a lateral subdivision of the "granular cove" (p. 604) (superior temporal plane and the superior temporal gyrus). The medial subdivision, being very rich in callosal connections, might correspond to an MMF source. However, the topographic segregation of these subdivisions has not been confirmed by physiological methods. Without accompanying anatomical imaging, the estimated ECDs, representing the "centre of activity" of the neuronal population, give only information about the relative locations of the sources rather than about the contributing cytoarchitectonic areas. The obtained results can be explained by two functionally different and spatially separable neuronal populations in the supratemporal auditory cortex which might be active in early sensory analysis of simple tones and detection of minor pitch deviations. One of the two sources represents the supratemporal M100 generator, whereas the olher represents the neuronal population generating the MMF, which can be regarded as an evoked field correlated with an automatic "change detection" process. The spatial and temporal separation of the two sources seem to support and verify N~i~.t~inen's proposal on the basic processing of auditory stimuli (1990), especially one of his two suggested processing modes: "task-independent" sensory analysis which is composed of a transient-detector and a "per-
547
manent" feature-detector system. The transient-detector system is organized in a stimulus-specific manner, structured according to tonotopic, amplitopic (Pantev et al. 1988, 1989a,b) or other feature maps, and activates the Nl-generator process. The result of this sensory analysis is stored in the short-term acoustic memory and reinforced by repetitions of the same stimulus feature to which the different stimuli are compared. Our data clearly demonstrate that, whereas slightly deviant frequencies do not activate different M100 generators, small deviations are already sufficient to trigger a spatially distinct mismatch generator. The authors thank Drs. R. N~i~it~inen and S. Makeig for thoroughly reviewing an earlier version of this manuscript.
References Annett, M. The binomial distribution of right, mixed and left-handed. Quart. J. Exp. Psychol., 1967, 19: 327-333. Cs6pe, V., Karmos, G. and Molnfir, M. Evoked potential correlates of stimulus deviance during wakefulness and sleep in cat animal model of mismatch negativity. Electroenceph. clin. Neurophysiol., 1987, 66: 571-578. Elberling, C., Bak, C., Kofoed, B., Lebech, J. and Saermark, K. Auditory magnetic fields from the human brain. Scand. Audiol., 1980, 9: 185-190. Elberling, C., Bak, C., Kofoed, B., Lebech, J. and Saermark, K. Auditory magnetic field. Auditory magnetic field from the human cortex: influence of stimulus intensity. Scand. Audiol., 1981, 10: 203 -207. Ford, J.M., Roth, W.T. and Kopell, B.S. Auditory evoked potentials to unpredictable shifts in pitch. Psychophysiology, 1976a, 13: 32-39. Ford, J.M., Roth, W.T. and Kopell, B.S. Attention effects on auditory evoked potentials to infrequent events. Biol. Psychol., 1976b, 4: 65-77. Galaburda, A. and Sanides, F. Cytoarchitectonic organization of the human auditory cortex. J. Comp. Neurol., 1980, 190: 597-61(/. Geselowitz, D.B. On the magnetic field generated outside an inhomogeneous volume conductor by internal current sources. IEEE Trans. Magn., 1970, 6: 346-347. Giard, M.H., Perrin, P. and Pernier, J. Brain generators implicated in processing of auditory stimulus deviance: a topographic ERP study. Psychophysiology, 1990, 27: 627-640. Hari, R. The neuromagnetic method in the study of the human auditory cortex. In: F. Grandori, M. Hoke and G.L. Romani (Eds.), Auditory Evoked Magnetic Fields and Electric Potentials. Adv. Audiol., Vol. 6. Karger, Basel, 1990: 222-282. Hari, R., H~im~il~iinen, M., Ilmoniemi, R., Kaukoranta, E., Reinikainen, K., Salminen, J., Alho, K., N~i~it~inen, R. and Sams, M. Responses of the primary auditory cortex to pitch changes in a sequence of tone pips: neuromagnetic recordings in man. Neurosci. Len., 1984, 50: 127-132. Hari, R., Pelizzone, M., M~ikel~i, 1.P., H~illstr6m, J., Leinonen, L. and Lounasmaa, O.V. Neuromagnetic responses of the human auditory cortex to on- and off-sets of noise bursts. Audiology, 1987, 26: 31-43. Hoke, M. Auditory evoked magnetic fields. In: E. Ba~ar (Ed.), Dynamics of Cognitive and Sensory Processing in the Brain. Springer, Berlin, 1988:311-318. Kaukoranta, E., Sams, M., Hark R., H~im~il~iinen, M. and N~i~it~inen,
548 R. Reactions of human auditory cortex to changes in tone duration. Hearing Res., 1989, 41: 15-22. Lounasmaa, O., Hari, R., Joutsiniemi, S.L. and HiimS_liiinen, M. MultiSQUID recordings of human cerebral magnetic fields may give information about memory processes. Europhys. Lett., 1989, 9: 6113-6//8. N~iS.tiinen, R. The role of attention in auditory information processing as revealed by event-related potentials and other brain measures of cognitive function, Behav. Brain Sci., 1990, 13: 201-288. N~iiitS.nen, R. and Galliard, A.W.K. The N2 deflection of ERP and the orienting reflex. In: A.W.K. Gaillard and W. Ritter (Eds.), EEG Correlates of Information Processing: Theoretical Issues. Elsevier, Amsterdam, 1983: 119-141. N~iS_t~inen, R. and Picton, T.W. The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure. Psychophysiology, 1987, 24: 375-425. N~i~it~inen, R., Galliard, A.W.K. and M~intysalo, S. Early selective attention effect on evoked potential reinterpreted. Acta Psychol. (Amst.), 1978, 42: 313-329. Niiiit~inen, R., Paavilainen, P., Alho, K., Reinikainen, K. and Sams, M. The mismatch negativity to intensity changes in an auditory stimulus sequence, ln: R. Johnson, Jr., J.W, Rohrbaugh and R. Pa,asuraman (Eds.), Current Trends in Event-Related Potential Research. Electroenceph. clin. Neurophysiol., Suppl. 40. Elsevier, Amsterdam, 1987: 125-131. Novak, G.P., Ritter, W., Vaughan, Jr., H.G. and Wiznitzer, M.L. Differentiation of negative event-related potentials in an auditory discrimination task. Electroenceph. clin. Neurophysiol., 1990, 75: 255-275. Paavilainen, P, Alho, K., Reinikainen, K., Sams, M. and Nii~itS_nen, R. Right hemisphere dominance of different mismatch negativities. Electroenceph. clin. Neurophysiol., 1991, 78: 466-479. Pantev, C., Liitkenh6ner, B., Hoke, M. and Lehnertz, K. Comparison between simultaneously recorded auditory-evoked magnetic fields and potentials elicited by ipsilateral, contralateral and binaural tone-burst stimulation. Audiology, 1986, 25: 54-61. Pantev, C., Hoke, M, Lehnertz, K., Liitkenh6ner, B., Anogianakis, G. and Wittkowski, W. Tonotopic organization of the human auditory cortex revealed by transient auditory evoked magnetic fields. Electroenceph. clin. Neurophysiol., 1988, 69: 160-170.
V. CSI~PE ET AL. Pantev, C., Hoke, M., Liitkenh6ner, B. and Lehnertz, K. Tonotopic organization of the auditory cortex: pitch versus frequency representation. Science, 1989a, 246: 486-488. Pantev, C., Hoke, M., Lehnertz, K. and Liitkenh6ner, B. Neuromagnetic evidence of an amplitopic organization of the human cortex. Electroenceph. clin. Neurophysiol., 1989b, 72: 225-231. Pantev, C., Gallen, C., ftampson, S,, Buchanan, S. and Sobel, D. Reproducibility and validity of neuromagnetic source localization using a large array biomagnetometer. Am. J. EEG Technol., 1991, 31: 83-11/1. Reite, M., Zimmerman, J.T. and Zimmerman, J.E. Magnetic auditory evoked fields: interhemispheric asymmetry. Electroenceph. clin. Neurophysiol., 1981, 5[: 388-392. Sams, M., Alho, K. and N~iS_t~inen,R. Sequential effects on the ERP in discriminating two stimuli. Biol. Psychol., 1983, 17: 41-58. Sams, M., Alho, K. and Nfi'atfinen, R. Short-term habituation and dishabituation of the mismatch negativity of the ERP. Psychophysiology, 1984, 21: 434-441. Sams, M., H~imS_l/iinen, M., Antervo, A., Kaukoranta, E.. Reinikainen, K. and Hari, R. Cerebral neuromagnetic responses ew)ked by short auditory stimuli. Electroenceph. clin. Neurophysiol., 1985, 61: 254-266. Sams, M., Kaukoranta, E., H~imiiliiinen, M. and NS_iit~inen, R. Cortical activity elicited by changes in auditory stimuli: different sources for the magnetic N1011m and mismatch responses. Psychophysiology, 1991, 28: 21-29. Sarvas, J. Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Phys. Med. Biol., 1987, 32: 11-22. Scherg, M., Vajsar, J. and Picton, T.W. A source analysis of the human auditory evoked potentials. J. Cogn. Neurosci., 1989, 1: 336-355. Simson, R., Vaughan, Jr., H.G. and Ritter, W. The scalp topography of potentials in auditory and visual discrimination tasks. Elcctroenceph, clin. Neurophysiol., 1977, 42: 528-535. Squires, K.C., Donchin, E., Herning, R.I. and McCarthy, G. On the influence of task relevance and stimulus probability on event-related stimulus components. Electroenceph. clin. Neurophysiol., 1977, 42: 1-14.