Lesions of Frontal Cortex Diminish the Auditory Mismatch Negativity

Electroencephalography and clinical Neurophysiology , 91 (1994) 353-362 353 © 1994 Elsevier Science Ireland Ltd. 0013-4694/94/$07.00

EEG92158

Lesions of frontal cortex diminish the auditory mismatch negativity

K. Alho b, D.L. Woods a A. Algazi a, R.T. Knight a and R. N i it inen b a Department o f Neurology and Neurobiology Center, UC Dacis, VA Medical Center, Martinez, CA 94553 (USA), and

b Cognitive Psychophysiology Research Unit, Department of Psychology, FIN-O0014 University o f Helsinki, Helsinki (Finland)

(Accepted for publication: 3 June 1994)

Summary Event-related brain potentials to non-attended auditory stimuli were recorded from patients with dorsolateral prefrontal cortex (DPFCx) lesions and from age-matched control subjects as they performed a visual reaction time task. Auditory stimuli consisted of monaural sequences of repetitive standard tones (I000 Hz) and occasional deviant tones of a higher frequency (1300 Hz). In comparison with control subjects, DPFCx patients showed enhanced P1 amplitudes (mean peak latency 50 msec), consistent with reduced frontally mediated gating of sensory input to the auditory cortex. The mismatch negativity (MMN) elicited by deviant tones was reduced in DPFCx patients over a broad latency range (130-210 msec), especially over the lesioned hemisphere and for tones delivered to the ear ipsilateral to the lesion. The results suggest that DPFCx and DPFCx-temporal projections play a critical role in involuntary orienting to physical changes in sequences of non-attended auditory stimuli.

Key words: Attention; Auditory; Event-related potentials; Frontal; Cortex; Orienting; Hemisphere; Evoked potential; Mismatch negativity; Visual

Physically deviant auditory stimuli occurring in a sequence of repetitive standard stimuli elicit a m/s- match negativity (MMN) in the auditory event-related brain potential (ERP; for a review, see N~i~it~inen 1992). The MMN typically overlaps with the N1 (peak latency 90-120 msec) and subsequent P2 (180-200 msec) deflections. Since N1 and P2 waves are also elicited by frequent standard stimuli, the MMN is best seen in difference waves obtained by subtracting standard-tone ERPs from deviant-tone ERPs.

The MMN has a fronto-central scalp distribution that can be modeled with intracranial sources in ante- rior regions of auditory cortex (Alho et al. 1986; Scherg et al. 1989; Novak et al. 1990; Paavilainen et al. 1991; Woods et al. 1993). An auditory cortex source is also suggested by magnetic (MEG) recordings in humans (Hari et al. 1984; Sams et al. 1985a, 1991; Kaukoranta et al. 1989; Lounasmaa et al. 1989; Tiitinen et al. 1993) and by intracranial recordings in the monkey (Javitt et al. 1991).

Correspondence to: David L. Woods, Ph.D., Chief, Clinical Neu- rophysiological Laboratory, Neurology Service (127), Northern Cali- fornia System of Clinics (NCSC), 150 Muir Road, Martinez, CA 94553 (USA). Tel.: (510) 3722571; Fax: (510) 2294897; E-mail: dlwoods@ ucdavis.edu.

N~iat~inen and collaborators (e.g., N~i~it~inen et al. 1989; N~i~it~inen 1990, 1992) have proposed that the MMN is generated by a mismatch between the physical features of deviant stimuli and a neuronal sensory- memory trace ("neuronal model"; Sokolov 1975) pro- duced by repetitive standard stimuli. A complete rep- resentation of the physical features of a repetitive stimulus is presumably stored in this memory trace, since a change in any feature (frequency, intensity, duration, location, etc.) elicits an MMN (N~i~it~inen 1992). The MMN is elicited by physical stimulus changes even when the subject's attention is directed to other auditory (e.g., N~i[it[inen et al. 1978; Alho et al. 1989) or visual stimuli (Sams et al. 1985b; Alho et al. 1992; Woods et al. 1992). Therefore, the MMN seems to be associated with the automatic analysis of stimulus features (N~i~it~inen 1992). However, the mismatch process that generates the MMN might be the initial stage in a chain of brain processes that leads to involuntary orienting of attention to changes in the acoustic environment (N~it~inen 1979; Sams et al. 1985b; N~i~it~inen and Picton 1987; Lyytinen et al. 1992).

Using scalp current density analysis, Giard et al. (1990) identified an additional MMN generator in frontal cortex. This frontal subcomponent of the MMN might be associated with the involuntary orienting of attention following the automatic stimulus-change de-

SSDI 0 0 1 3 - 4 6 9 4 ( 9 4 ) 0 0 1 7 3 - 1

354 K. ALHO ET AL.

FRONTAL

2 3

7'

~ : _ - ,L ~ .~ .~_-~ . '~%

4 5 6 7

Fig. I. Reconstructed frontal lesions in the patients (N = 10) studied. Right hemisphere lesions (N = 3) have been reflected onto the lett hemisphere for display. Horizontal lines indicate the planes of section. The shading reflects the percentage of patients with damage.

tection mechanism indexed by the auditory cortex subcomponent of the MMN (Giard et al. 1990; N~ifit~inen 1992). However, it has been difficult to disentangle the relative contributions of frontal and auditory cortex to MMN generation because of temporal and distribu- tional overlap (Giard et al. 1990). While the frontal subcomponent would be expected to have an amplitude maximum at frontal scalp sites, the auditory cortex subcomponent of the MMN might also be largest at frontal sites because of generator orientation (Scherg et al. 1989; Tiitinen et al. 1993).

One frontal structure that might be involved in MMN generation is the dorsolateral prefrontal cortex (DPFCx) which is known to have an important role in the control of attention (Knight et al. 1981; Woods and Knight 1986; Fuster 1989; Knight 1991). The aim of the present study was to evaluate the role of this cortical area in MMN generation by comparing MMNs in patients with unilateral DPFCx lesions and in healthy control subjects.

M e t h o d s

Subjects E R P s w e r e r e c o r d e d in 10 p a t i e n t s (ma le s , 6 0 - 7 5

yea r s ) w i t h u n i l a t e r a l D P F C x l e s ions (7 l e f t - s i d e d , m e a n

l e s ion v o l u m e 41.0 ml; Fig. 1) d u e to i n f a r c t i o n s o f t h e

precentral branch of the middle cerebral artery. Pa- tient data were compared with data from 13 healthy, age-matched control subjects (9 males, 4 females, 54-80 years). The hearing thresholds of all subjects were tested before the experiment. 1 All subjects had a normal corrected visual acuity and no history of alco- holism, drug abuse or psychiatric problems. All subjects were paid volunteers who participated after giving informed consent according to institutional guidelines.

t Before the experiment, each subject underwent audiometric testing at 125, 250, 500, 1000, 1500, 2000, 4000, and 8000 Hz. At the highest frequencies (2000 Hz or higher), most control subjects showed age- related elevations of thresholds. However, at frequencies below 2000 Hz, their thresholds were always lower than 25 dB SPL. The patients showed somewhat greater hearing losses than the controls. At 1000 Hz, the thresholds of patients varied between 15 and 45 dB SPL (5 patients had thresholds higher than 25 dB SPL). At 1500 Hz, the thresholds varied between 10 and 55 Hz SPL (6 patients had thresholds higher than 25 dB SPL). In the patients, the mean hearing thresholds for 1000 Hz in the ear contralateral and ipsilateral to the lesioned hemisphere were 28 and 25 dB SPL, respectively. For 1500 Hz, these thresholds were 33 and 29 dB SPL, respectively. Although the patients had, on the average, higher thresholds than the c o n t r o l

subjects, the 1000 Hz and 1300 Hz tones of 82 dB SPL used in the present ERP experiment were well above the hearing thresholds for all patients. Furthermore, an analysis of variance including Fre- quency (1000 Hz and 1500 Hz) and Ear (contralateral vs. ipsilateral to the lesion) as factors showed no significant differences in the thresholds of patients between the two frequencies or two ears.

F R O N T A L LESIONS A N D T HE M M N 355

Stimuli and procedure Auditory and visual stimuli were presented in a

random order in blocks ,of 440 stimuli. Interstimulus intervals (ISis) varied randomly between 200 and 400 msec in 20 msec increments. Auditory stimuli were pure tones with a duration of 50 msec (5 msec rise and fall times) and intensities of 82 dB SPL. They were presented monaurally through TDH-39 headphones over a continuous binaural broadband masking noise (60 dB SPL). Standard tones of 1000 Hz occurred with a probability of 85% and deviant tones of 1300 Hz with a probability of 10%. In half of the stimulus blocks, the tones were delivered to the left ear and in the other half to the right ear.

In addition to auditory stimuli, 5% of stimuli were visual targets (occurring with a mean inter-target inter- val of 6 sec). Visual targets were white vertical gratings on a black background (spatial frequency 2 c / d e g ; luminance 6 fL; contrast 0.99; size 3.9 ° x 3.9°; duration 50 msec). They appeared in the center of a video monitor subtending l l . 2 ° x 15.4 ° of visual angle and positioned 1.5 m in front of the subject.

The experiment consisted of 12-16 stimulus blocks. The left-and right-ear blocks were delivered in an alternating order. During each stimulus block, the subject was instructed to ignore the tones, to focus on a small fixation star in the center of the television screen, and to respond with a button press of the right-hand thumb to each visual stimulus. The subjects were instructed to be as fast as possible in their responses. Fixation was assured in the beginning of each stimulus block and eye position was continuously monitored with vertical and horizontal EOG (see below) and through a closed-circuit video camera.

Performance in the visual detection task was scored by computer. Responses occurring from 80 to 1000 msec after target onset were defined as hits and responses given outside this window were defined as false alarms.

After the main experiment, subjects participated in a supplementary auditory discrimination task without ERP recording. In this task, auditory stimuli were similar to those used in the main experiment, but the subjects' task was to press the response button, as fast and accurately as possible, after each 1300 Hz deviant tone occurring among standard tones (1000 Hz). Re- sponses given within 80-1000 msec from a deviant-tone onset were defined as hits and other responses were defined as false alarms. Two left-ear stimulus blocks and 2 right-ear blocks were presented. The order of the blocks was counterbalanced between subjects.

EEG recording and averaging The EEG (bandpass 0.1-100 Hz) was continuously

sampled (256 Hz / channe l ) from 27 scalp electrodes: Fpl , Fpz, Fp2, F7, F3, Fz, F4, F8, PAl (left preauricu-

lar point), T1 (between F7 and T3), T3, C3, Cz, C4, T4, T2 (between F8 and T4), PA2 (right preauricular point), M1 (left mastoid), T5, P3, Pz, P4, T6, M2 (right mastoid), O1, Oz, and 02. Additional electrodes were attached below the left eye (BE) and lateral to the left eye (LE) to monitor EO G artifacts. All electrodes were referenced to interconnected, ECG-balanced electrodes at the base of the neck (Woods and Clayworth 1985).

E E G epochs starting 200 msec before stimulus onset were averaged off-line by computer. Trials contaminated by vertical or horizontal eye movements (peak- to-peak amplitude at Fpz, BE, or LE exceeding 80 /~V), amplifier clipping, bursts of EMG activity, or other artifacts were automatically excluded from the averages. The first stimulus was omitted from the average in each stimulus block. Only hit trials were in- cluded in the ERPs to visual target stimuli. To avoid contamination of auditory ERPs at posterior scalp sites by ERPs to preceding visual stimuli (Alho et al. 1992), ERPs to auditory stimuli occurring within 600 msec of a visual stimulus were excluded from the average.

Data analysis After averaging, ERPs were digitally filtered to

eliminate frequencies above 40 Hz. Difference waves were obtained by subtracting standard-tone ERPs from deviant-tone ERPs separately for each ear of stimulation. ERP peak amplitudes and latencies were measured in relation to stimulus onset and ERP mean voltages were measured over consecutive 20 msec intervals between 10 and 510 msec (10-30 msec, 30-50 msec, 50-70 msec, etc.) in reference to a 200 msec prestimulus baseline. 2

The results were statistically evaluated with analyses of variance for repeated measures. Significance levels were adjusted with the Greenhouse-Geisser correction when appropriate. However, the original degrees of freedom are reported for each analysis.

2 Due to the short ISis (random 200-400 msec) used in the present experiment, ERPs may have been contaminated by overlapping t ime-smeared ERPs to preceding and subsequent stimuli. Al though a contamination of auditory ERPs by the large P3 deflection elicited by visual target stimuli (see Fig. 6) was avoided by discarding ERPs to tones immediately following visual targets (see Methods), ERPs to tones were overlapped and affected by t ime-smeared ERPs to preceding and subsequent auditory stimuli, as well as by subsequent visual stimuli (for a fur ther discussion, see Woldorff 1993). However, this response overlap would affect similarly the ERPs to deviant and standard tones delivered in a random order. Therefore, overlapping activity would not affect the mismatch negativity (MMN), the main topic of the present study, which was determined as a difference between the ERPs to deviant and standard tones. In these difference waves, the effects of overlapping responses are canceled by subtrac- tion.

356 K. ALHO ET AL.

In the analysis of performance, the percentages of hits and false a larms/delayed responses, and mean reaction times (RTs) were calculated for the visual task of the main experiment and for the supplementary auditory discrimination task. Analyses of variance were used to compare RTs and performance in the patients and controls.

Results

Performance The mean RT to visual targets was 405 msec and

346 msec for the patients and controls, respectively (F (1, 21) = 7.58, P < 0.02). Hit rates did not differ significantly (controls, 94% vs. patients, 88.4%; F (1, 21)= 1.59, NS), nor were significant differences found between patients and controls in the percentages of false a larms/delayed responses.

In the supplementary auditory discrimination test, the patients detected fewer deviant tones (77.4%) than the controls (95.9%; F (1, 21)= 6.92, P < 0.02), and had longer RTs (439 msec vs. 375 msec, F (1, 21) = 7.40, P < 0.02). In the patients, 15.6% of the responses were false alarms, while in the controls, this percentage was 13.7%. In the patients, discrimination performance tended to be less accurate for deviant tones delivered to the ear ipsilateral to the lesioned hemisphere, but ear effects on RT and detection performance did not reach significance.

Auditory ERPs Grand-average ERPs to standard tones in the

DPFCx patients and control subjects are shown in Fig. 2. In both groups, standard tones elicited N1 and P2 deflections with maximal amplitudes at central and frontal scalp sites. Table I shows mean peak amplitudes and latencies at Cz for the N1 and P2 deflections elicited by standard tones. The N1 and P2 deflections had similar peak amplitudes for the patients and controls. Although the N1 peak latencies tended to be shorter, and the P2 peak latencies longer in the patients than in the controls, these latency differences failed to reach significance.

As seen in Fig. 2 (insert) the amplitude of the P1 deflection was relatively enhanced in the patients. An analysis of variance for P1 amplitudes (measured at Fz as mean voltages over 30-50 msec from stimulus onset) in the standard-tone and deviant-tone ERPs indicated a significant difference (on the average 0.4 /zV) between the two groups (F (1, 21) = 4.95, P < 0.05). For the standard tones, the early portion of the N1 deflection (before the N1 peak) also tended to be enhanced in the patients in relation to the controls (see Fig. 2), but this effect, tested at 50-70 msec, 70-90 msec, and 90-110 msec was not statistically significant.

. . . . . . . . . . . - --rV- ~ , ~

/ ~ J ] i ~c ~ ~ I

CONTaOT.S :~vv_~_ ~ " r ' ~ - ...... FRONTALS

. , , %

/ i f ~ { i ii= ii~'e~e@iiiiiflflif ~c, ~ N ~ 1 "

CONTaOLS ' ~ ~v~___.= r . . . . . - ...... FRONTALS

Fig. 2. Grand-average ERPs to standard tones from control subjects and DPFCx (frontal) patients. ERPs were transposed in patients so that ERPs on left of the figure were always ipsilateral to the lesion and electrodes have been relabeled accordingly (i = ipsilateral to lesion, c = contralateral). Top: ERPs to tones ipsilateral to the lesioned hemisphere (left ear for the control subjects). The arrow shows the ear of stimulation. Bottom: ERPs to contralateral tones. Inserts: enlarged ERPs from Fi and Fc electrodes. Lesions have been

outlined schematically.

TABLE I

Mean N1 and P2 peak amplitudes and latencies (S.E.M. in parenthe- ses) in patients with frontal lesions (N = 10) for standard tones delivered to the ear contralateral (right ear in 7 patients) or ipsilateral to the lesion, and in control subjects (N = 13) for standard tones delivered to the right or left ear.

Ear N1 and P2 amplitudes

Frontals ( N = 10) Controls ( N = 13 )

(#V) (msec) (/zV) (msec)

N1 Cont ra / r igh t - 1 . 2 ( ! 0 . 2 ) 114 (+8 ) -1 .2(_+0.6) 127(_+7) NI lps i / le f t -1 .3(_+0.2) 107(_+9) -1 .2(_+0.7) 120(_+8) P2 Can t ra / r igh t 1.1(_+0.2) 223(_+_6) 1.1(_+0.5) 219(_+7) P2 Ips i / lef t 1.2(_+0.1) 225(_+7) 1.2(_+0.5) 214(_+_4)

FRONTAL LESIONS AND THE MMN 357

At latencies longer than 100 msec, E R P negativity to deviant tones delivered to the ear ipsilateral to the lesion (left ear in 7 patients) was reduced in the patients in relation to controls (Fig. 3). A smaller reduction was seen for deviant tones delivered to the ear contralateral to the lesion. As discussed below, these effects appeared to be caused by smaller M M N s in the DPFCk patients.

As seen in compar isons of Figs. 2 and 3, the N1 deflection was larger in ampli tude and longer in dura- t ion to deviant than s tandard tones. The early par t of the negativity to deviant stimuli may have received some contr ibut ion f rom enhanced N1 components which may have been less refractory for the infrequently occurr ing deviant tones (see, e.g., Scherg et al. 1989; Lang et al. 1990; Woods 1994), while the later port ion appeared to reflect mainly the addit ion of an

\~a~ . ~ _ W~:..I" AUDITORY

~ ~ - - C O N T R O L S --,"s-~..~ : ~ . . . > r ' - ' - " . . . . . . . FRONTALS

~ 1 ,-~N~ 1 I E _ t / / \ ~.v,.. I

AUDITORY

" '-"< ....... FRONTALS

Fig. 3. Grand-average ERPs to deviant tones for control subjects (solid lines) and DPFCx (frontal) patients (broken lines). Top: ERPs to tones ipsilateral to the lesioned hemisphere (left ear for the control subjects). Bottom: ERPs to contralateral tones. The arrow shows the ear of stimulation. Inserts: enlarged ERPs from Fi and Fc

electrodes. Lesions have been outlined schematically.

/~ I t-.MMN ~ Fi

~ ~---.~--/ DIFFERENCE . . . . . ~ ~ / IDEVS-STANDS

~ ~ CONTROLS + ~ , . . --1,vt 5 - ~ ' - - . . . . . . FRONTALS

tt ..... /AN ~- I MMN I

~" ~ ~ / DEVS-STANDS

o.rROLS

Fig. 4. Difference waves obtained by subtracting grand-average ERPs to standard tones from grand-average ERPs to deviant tones for DPFCx (frontal) patients and control subjects. Top: difference waves for tones ipsilateral to the lesioned hemisphere. Bottom: difference waves for contralateral tones. Inserts: enlarged ERPs from Fi and Fc

electrodes. Lesions have been outlined schematically.

MMN. The M M N was more readily seen in difference waves obta ined by subtract ing the E R P s to s tandard tones f rom the E R P s to deviant tones (Fig. 4). It was largest at frontal scalp sites for both subject groups. Difference waves showed a significant negativity at 70-210 msec (consecutive 20 msec periods at Fz, F (1, 21) ranged f rom 12.18 to 152.03, P < 0.005 in all cases).

M M N s were significantly reduced in DPFCx patients in compar ison with control subjects (Fig. 4): significant Group effects were observed at Fz at 130- 150 msec (controls = - 2 . 2 /zV, pat ients = - 1 . 3 / x V ; F (1, 2 1 ) = 6.99, P < 0.015) and at 150-170 msec (controls = - 2.0/xV, patients = - 1.0/~V; F (1, 21) = 5.79, P < 0.025). Similar Group effects were observed in analyses of variance of E R P ampli tudes at the frontal Fi and Fc electrodes located over the lesioned and

358 K. ALHO ET AL.

intact hemispheres, respectively (Fig. 4; 130-150 msec, F (1, 21) = 7.88, P < 0.010; 150-170 msec, F (1, 21) = 7.33, P < 0.015).

Analyses of MMN amplitudes at Fi and Fc electrodes showed significant Group x Ear x Hemisphere interactions at 170-190 msec (F (1, 21)--6.84, P < 0.015) and 190-210 msec (F (1, 21)= 5.50, P < 0.03, with a trend at 150-170 msec (F (1, 21)= 3.57, P < 0.07). This was due to the fact that MMN amplitudes were more reduced over lesioned than intact hemispheres for tones delivered to the ear ipsilateral to the lesion than for contralateral tones (Fig. 4). For example, at 170-190 msec, mean MMN amplitudes in the patients for stimuli delivered ipsilaterally to the lesion (left ear in 7 patients) were - 0 . 2 /zV and -0.6/_tV at Fi and Fc, respectively, whereas in the controls, corre- sponding MMN amplitudes were - 1.5 /_iV and .... 1.3 ~tV, respectively.

Fig. 5 shows the MMN scalp distributions in the patient and control groups. In the controls, the MMNs to left- and right-ear deviant tones had frontal distributions that were symmetrical over the two hemispheres. In the DPFCx patients, the MMN to deviant tones delivered to the ear ipsilateral to the lesion had a frontal distribution and was reduced over the lesioned hemisphere (Fig. 5, see statistical analysis above). The MMN to deviant tones delivered to the ear contralateral to the lesion showed a tendency toward a similar ipsilesional reduction (at 130-170 msec; see Fig. 5) which failed to reach significance.

It is important to note that the reduced MMN in DPFCx patients retained its frontal scalp distribution.

lpsilateral Ear Con~alateral Ear

Controls Frontals Controls Frontals

120 m s - ~ 1 @ , ' 1 ~ @ @ ' @ "

@.@.

180 m s - i , . @ 4 , . @ @ @ "

Fig. 5. Normalized MMN scalp distributions over consecutive 20 msec measuring windows (centered at the time shown) in DPFCx (frontal) patients and control subjects. In the patients, the left side of the head represents the lesioned hemisphere. Left: MMN distribution for deviant stimuli delivered to the ear ipsilateral to the lesion (left ear for the controls). Right: MMN distribution for deviant stimuli delivered to the contralateral ear. Solid lines = negative volt-

ages; dashed lines = positive voltages.

TABLE 11

Mean MMN amplitudes in control subjects (N = 13) and in patients with frontal lesions (N = 10). Mean voltage measures from difference waves for standard tones delivered to the ear contralateral (right ear in 7 patients) or ipsilateral to the lesion, and to standard tones delivered to the right or left ear in controls.

MMN amplitude (deviant- standard) over 150-170 msec

Contra ear (right) Ipsi ear (left)

Controls Frontals Controls Frontals

Fpi - 1.8 .... 1.2 - 1,3 -- [).8 Fpz - 1.7 1.2 - 1.3 --0.7 Fpc -1 .7 1.1 - 1.1 -0.6

Fi - 1.9 - 1.4 - 2.(I .... 0.5 Fz -- 1,9 • 1.2 -2.1 (1,7 Fc -. 1.7 1,2 2.1 .... I.t)

Ci - 1.2 - I.II 1.3 .... 0.5 Cz - 1,2 - I.I -- t.6 (L5 ('c -1.0 12 1.7 -(I.8

Pi + 0 .1 ~i.5 + 0 . 4 - (L6 Pz - 0,2 0.4 0.5 + 0.3 Pc (12 0.4 -- 0.6 + 0.2

O i + 0.4 0.2 + O. l 4- (1.7

Oz + 0.3 - 0, 1 - 0.1 + (1.6 Oc + 0.4 - 0.3 - 0.2 + 0.3

In particular, DPFCx lesions did not produce a focal frontal amplitude reduction: the MMN reduction in patients was similar (in percentage terms) at frontal, central and parietal electrodes over the lesioned hemisphere (Table II).

The MMN was followed by a smaller, frontally distributed late negative shift in deviant-tone ERPs (Fig. 4). Analyses of variance for the mean ERP amplitudes indicated a significant difference between ERPs to deviant and standard tones at 290-510 msec at frontal sites (F (1, 21) ranged from 4.78 to 38.51, P < 0.05 in all cases). The late negativity tended to be reduced in the DPFCx patients over the lesioned hemisphere (see Fig. 4), but this effect failed to reach significance.

ERPs to visual target stimuli ERPs to the infrequently occurring visual target

stimuli are shown in Fig. 6. These stimuli elicited an occipito-parietal NI preceded by a short-duration P1. No significant group differences were noted at P1 latency. Analyses of variance for N1 amplitude at O1, Oz, and 0 2 indicated that there was no significant difference in peak amplitude between the groups (patients -4 .0 /~V; controls -5.4/. tV). However, N1 peak latencies were significantly shorter~in patients than in controls (patients 129 msec vs. controls 148 msec; F (1, 21) = 4.33, P < 0.05).

Visual targets also elicited a P3 which was largest at Pz (peak amplitude: patients 11.7 p.V; controls 13.3/.tV; peak latency: patients 391 msec; controls 375 msec).


f I n , I

F° , " PS.[

CONTROLS . . . . . . .

Fig. 6. Grand average ERPs to visual target stimuli for control subjects and for DPFCx (frontal) patients. Insert: enlarged ERPs from Fi and Fc electrodes. Lesions have been outlined schematically.

Differences in the P3 peak amplitude and latency between the two groups did not reach significance.

The P3 was preceded by an N2 wave peaking at about 250 msec from stimulus onset (Fig. 6). Visual N2 amplitudes were maximal over posterior temporal electrodes suggesting a visual cortex or visual association cortex generator. In control subjects, the N2 was also larger over the left hemisphere (see electrodes Fc and Cc in Fig. 6), i.e., the hemisphere contralateral to the hand used for responding consistent with a possible contribution from the time-smeared movement-related potentials (MRPs) preceding the button press.

The N2 was reduced in amplitude over the lesioned hemisphere in the frontal patients. For example, at 230-250 msec, the mean visual ERP amplitudes for the patients at Fi and Fc were 5.0/zV and 3.5 ~V respectively, while in the control subjects F3 and F4 mean amplitudes were 0.2 ~V and 2.0 /xV, respectively. Analyses of variance (factors: Group, Hemisphere) for mean visual ERP amplitudes (measured over consecutive 20 msec periods) at Fi and Fc indicated significant Group × Hemisphere interactions at 210-290 msec from stimulus onset (F (1, 21) = 4.42-7.71, P < 0.05 in all cases). Similar significant Group × Hemisphere interactions were also observed for the Ci and Cc amplitudes at 210-290 msec (F (1, 21) = 4.52-6.73, P < 0.05 in all cases). A strong tendency toward N2 reductions was also seen at temporal, posterior temporal, and parietal electrodes over the lesioned hemisphere.

Discussion

Frontal lesions and sensory ERPs Auditory stimuli elicited sensory ERPs with a simi-

lar component structure in frontal patients and control

subjects. P1 amplitudes were larger in the patients than in the controls, both to standard and deviant tones. An enhancement of middle latency auditory evoked potentials in patients with unilateral DPFCx lesions has been previously reported by Knight et al. (1989). They suggested that this early ERP enhancement was associated with reduced DPFCx mediated thalamic gating of sensory inputs to the auditory cortex (see also Skinner and Yingling 1977; Yamaguchi and Knight 1990; Knight 1991). In the present study, the occipital N1 deflection elicited by visual stimuli had shorter peak latencies in DPFCx patients than in controls, an effect that might also be caused by reduced thalamic gating of visual inputs.

Frontal lesions and the mismatch negativity MMN amplitudes were attenuated in the DPFCx

patients in comparison with the control subjects, consistent with the possible involvement of frontal cortex in the generation of MMN (N~i~it~inen 1992). Previous studies have shown that DPFCx lesions also cause decreased N2-P3a responses elicited by unexpected, novel auditory stimuli (Knight 1984, 1990, 1991; Scabini et al. 1989).

The present MMN attenuation in the patients can- not be explained by elevated hearing thresholds. Tone intensities were well above the hearing thresholds of both ears in all patients. Moreover, thresholds were similar for ipsilesional and contralesional ears in the patients, even though MMN abnormalities were more marked for tones presented to the ipsilesional ear. Although the performance in the supplementary auditory discrimination task was not as good in the patients as in the control subjects, all patients were able to discriminate deviant from standard stimuli. Several studies have suggested a relationship between discrimination performance and MMN amplitude a n d / o r latency (Sams et al. 1985b; Lang et al. 1990; Novak et al. 1992). Therefore it is possible that the less accurate discrimination of deviant tones and longer RTs to these tones in patients than in controls were partly caused by dysfunctioning of the stimulus change detection mechanism generating the MMN. However, visual RTs were also prolonged in the patients. This is consistent with the possibility that the prolongation in auditory RTs seen in the patients may have been partly caused by general problems in motor performance associated with DPFCx lesions (Singh and Knight 1990).

It might be argued that smaller MMNs to auditory deviants were elicited in patients because they found the visual task more difficult than the controls and so had less spare attentional capacity for processing irrelevant auditory stimuli (cf., N~i~it~inen 1991; Woldorff et al. 1991). However, previous results indicate that the amount of attention demanded by a visual task has no effect on the MMN to deviant stimuli in unattended

360 K. ALItO ET At..

auditory inputs (Alho et al. 1992). Moreover, attending to auditory stimuli does not enhance the MMN to large frequency changes such as those used in the present study (Alho et al. 1992). In any case, even differential between-group effects of attention on the MMN would not explain the group differences in MMN scalp distribution observed in the present study.

Previously, a contribution of a frontal source to the MMN was suggested by scalp current density maps in normal subjects (Giard et al. 1990). In the present study, the MMN was at tenuated over the lesioned hemisphere, which might be regarded as evidence that the lesions damaged a frontal MMN generator. How- ever, this interpretation is contradicted by the similar- ity in distribution over the lesioned and intact hemispheres: there was no alteration in the frontality of the MMN distribution. This result could be explained by a frontal MMN source in ventro-lateral frontal or ba- sofrontal regions. Alternatively, a DPFCx source may have been spared because most of the patients had a frontal lesion in the left hemisphere while the generator of the frontal MMN component may preferentially involve the right DPFCx cortex (Giard et al. 1990). The small number of frontal patients with right-hemisphere lesions did not permit an adequate evaluation of a possible right DPFCx generator.

Another explanation for MMN attenuation is that the lesions resulted in an attenuation of MMN generation in ipsilesional auditory cortex. Models of MMN sources show that generators in left and right auditory cortices cause a frontally dominant MMN scalp distribution even in the absence of a frontal MMN subcomponent (Scherg et al. 1989). The persistence of neuronal sensory-memory traces necessary for the mismatch process might depend on sustaining fronto-temporal projections to auditory cortex. There is also evidence that the MMN may have at least two auditory cortex subcomponents (Paavilainen et al. 1991). Per- haps the generator of the later one of these subcomponents is triggered by frontal activation caused by a stimulus change, and is therefore at tenuated by frontal lesions. Whichever is the explanation for the present MMN attenuation in the DPFCx patients, the lack of reduction in P1, N l, and P2 amplitudes in these patients suggests that any effect of frontal lesion on functioning of the auditory cortex would have specifi- cally affected discrimination or memory functions (Baddeley and Wilson 1988; Wilson et al. 1993) rather than auditory analysis in general.

The deteriorated performance of the DPFCx patients in the supplementary auditory discrimination task is consistent with dysfunctions of the change-detection mechanism associated with the MMN (NS.5.fii- hen 1992). However, the present lesions might have been expected to more markedly disrupt the processing of auditory stimuli in contralateral hemispace since

unilateral lesions of frontal cortex usually result in contralateral deficits in orienting (Heilman and Valen- stein 1972; Heilman and Watson 1977; De Renzi et al. 1989; Hugdahl et al. 1991). However, in some patients unilateral frontal lesion may be associated with inap- propriate orienting, rather than neglect, to contralesional stimuli ("visual grasp"; Butter et al. 1988) and impaired orienting to stimuli in ipsilateral hemispace (Kwon and Heilman 1991). In the present study, we found evidence for ipsilateral auditory neglect: the MMN attenuation seen in the DPFCx patients was more marked for ipsilateral than contralateral tones and frequency discrimination in the patients tended to be less accurate for ipsilesional tones. The ear-specific reduction in the MMN suggests that it did not result simply from a non-specific decline in arousal a n d / o r motivation. Rather, it suggests that separate cortical circuits are involved in processing auditory stimulus change in left and right auditory hemispace.

Another reason for the reduction in MMN following tones delivered ipsilaterally to the lesion may relate to circuitry of MMN generation. Scalp current density maps (Giard et al. 1990) suggest that MMN sources in the left and right auditory cortices are more strongly activated by stimuli delivered to the contralateral ear than by stimuli delivered to the ipsilateral ear. Hence, projections from the ipsilateral frontal lobe may pro- vide relatively more important inputs to the weaker MMN generator process activated by ipsilateral stimuli. For example, this might occur if the MMN generator in auditory cortex required convergent projections t!rom both primary auditory cortex and the ipsilateral frontal lobe.

The MMN was IoIIowcd by another negativity to deviant tones. A similar late negativity has been observed in previous studies (NS~it~inen et al. 1982; Alho et al. 1992). N~iS_t~inen ct al. (1982)suggested that this negativity folk)wing the MMN might reflect "'sensitization processes" after an occurrence of a stimulus change which elicits an MMN but fails to trigger any subsequent endogenous processes. Such "sensitization" might express automatic preparation for detecting possible subsequent stimulus changes. In the present study, no significant attenuation of the late negativity following the MMN was observed in the DPFCx patients suggesting different neural origins for the late negativity and the MMN.

Frontal lesions and ERPs to visual targets ERPs to visual target stimuli showed reduccd N2

amplitudes in the DPFCx patients. These effects were particularly pronounced over the left hemisphere suggesting a possible contribution from reduced movement-related cortical potentials (MRPs) that would have made t ime-smeared contributions in the stimulus- locked averages to targets (Singh and Knight 1990).


H o w e v e r , t h e t e n d e n c y t o w a r d N 2 r e d u c t i o n o v e r t e m -

p o r a l a n d p a r i e t a l c o r t e x ips i l a t e ra l to t h e D P F C x

l e s i o n s s u g g e s t s t h a t f r o n t o - p a r i e t a l p r o j e c t i o n s

( Y e t e r i a n and P a n d y a 1985; S e l e m o n a n d G o l d m a n -

R a k i c 1988) m a y be n e c e s s a r y fo r t h e g e n e r a t i o n o f a

v isual N 2 o f n o r m a l a m p l i t u d e ( D e a c o n e t al. 1991).

Conclusions T h e M M N a t t e n u a t i o n in p a t i e n t s w i th D P F C x le-

s ions sugges t s t ha t t h e f ron t a l c o r t e x plays a ro l e in t h e

n e u r a l c i r cu i t ry u n d e r l y i n g t h e M M N . T h e p r e s e n t

M M N a t t e n u a t i o n in t h e D P F C x p a t i e n t s was n o t

c a u s e d by a g e n e r a l r e d u c t i o n o f a u d i t o r y E R P a m p l i -

tudes . O n the con t r a ry , P1 a m p l i t u d e s w e r e e n h a n c e d

in t h e pa t i en t s . T h i s e f f ec t m a y h a v e b e e n c a u s e d by

r e d u c e d f ron t a l g a t i n g o f s enso ry i n p u t f r o m t h a l a m u s

to t h e a u d i t o r y cor tex . A l so , s h o r t e r v isual N1 p e a k

l a t e n c i e s in t he p a t i e n t s m a y h a v e i n d e x e d r e d u c e d

t h a l a m i c g a t i n g o f v i sua l inpu t .

D P F C x les ions a lso s l o w e d visual t a r g e t d e t e c t i o n

a n d r e d u c e d N 2 a m p l i t u d e s to t h e t a r g e t s t imul i . T h e s e

e f fec t s m i g h t be c a u s e d in pa r t by m o t o r i m p a i r m e n t

and r e d u c e d m o v e m e n t - r e l a t e d nega t iv i t i e s . A l t e r n a -

t ively, t hey m a y r e f l e c t d a m a g e to a f r o n t o - p a r i e t a l

n e t w o r k i nvo lved in t a r g e t d e t e c t i o n a n d d i s c r i m i n a -

t ion.

Supported by DC 01049, NS 32893, a US Public Health Service International Research Fellowship (F05 TW04283), the Academy of Finland, the University of Helsinki, and the VA Research Service.

References

Alho, K., Paavilainen, P., Reinikainen, K., Sams, M. and N~i~it~inen, R. Separability of different negative components of the event-related potential associated with auditory stimulus processing. Psy- chophysiology, 1986, 23: 613-623.

Alho, K., Sams, M., Paavilainen, P., Reinikainen, K. and Nii~it~inen, R. Event-related brain potentials reflecting processing of relevant and irrelevant stimuli during selective listening. Psychophysiology, 1989, 26: 514-528.

Alho, K., Woods, D.L., Algazi, A. and N~ifit~inen, R. Intermodal selective attention. II. Effects of attentional load on processing auditory and visual stimuli in central space. Electroenceph. clin. Neurophysiol., 1992, 82: 356-368.

Baddeley, A. and Wilson, B. Frontal amnesia and the dysexecutive syndrome. Brain Cogn., 1988, 7: 212-230.

Butter, C.M., Rapcsak, S., Watson, R.T. and Heilman, K.M. Changes in sensory inattention, directional motor neglect, and "release" of the fixation reflex following a unilateral frontal lesions: a case report. Neuropsychologia, 1988, 26: 533-545.

Deacon, D., Breton, F., Ritter, W. and Vaughan, Jr., H.G. The relationship between N2 and N400: scalp distribution, stimulus probability and task relevance. Psychophysiology, 1991, 28: 185- 200.

De Renzi, E., Gentilini, M. and Barbieri, C. Auditory neglect. J. Neurol. Neurosurg. Psychiat., 1989, 52: 613-617.

Fuster, J. The Prefrontal Cortex, 2nd Edn. Raven Press, New York, 1989.

Giard, M.H., Perrin, F., Pernier, J. and Bouchet, P. Brain generators implicated in the processing of auditory stimulus deviance: a topographic event-related potential study. Psychophysiology, 1990, 27: 627-640.

Hari, R., H~im~il~iinen, M., Ilmoniemi, R., Kaukoranta, E., Reini- kainen, K., Salminen, J. and Sams, M. Responses of the primary auditory cortex to pitch changes in a sequence of tone pips: neuromagnetic recordings in man. Neurosci. Lett., 1984, 50: 127- 132.

Heilman, K.M. and Valenstein, E. Auditory neglect in man. Arch. Neurol., 1972, 26: 32-35.

Heilman, K.M. and Watson, R.T. The neglect syndrome - a unilateral defect of the orienting response. In: S. Harnad, R.W. Doty, L. Goldstein, J. Jaynes and G. Krauthamer (Eds.), Lateralization in the Nervous System. Academic Press, New York, 1977: 285- 301.

Hugdahl, K., Wester, K. and Asbjornsen, A. Auditory neglect after right frontal lobe and right pulvinar thalamic lesions. Brain Lang., 1991, 41: 465-473.

Javitt, D.C., Schroeder, C.E., Arezzo, J. and Vaughan, Jr., H.C. Selective inhibition of "processing-contingent" auditory event-related potentials by phencyclidine (PCP)-like agent MK-801. Elec- troenceph, clin. Neurophysiol., 1991, 79: 65P.

Kaukoranta, E., Sams, M., Hari, R., H~im~il~iinen, M. and N~i~it~inen, R. Reactions of human auditory cortex to a change in tone duration. Hear. Res., 1989, 41: 15-21.

Knight, R.T. Decreased response to novel stimuli after prefrontal lesions in man. Electroenceph. clin. Neurophysiol., 1984, 59: 9-20.

Knight, R.T. Neural mechanisms of event related potentials: evidence from human lesion studies. In: J.W. Rohrbaugh, R. John- son and R. Parasuraman (Eds.), Event-Related Brain Potentials: Issues and Interdisciplinary Vantages. Oxford University Press, New York, 1990: 3-18.

Knight, R.T. Evoked potential studies of attention capacity in human frontal lobe lesions. In: H. Levin, H. Eisenberg and F. Benton (Eds.), Frontal Lobe Function and Dysfunction. Oxford Univer- sity Press, Oxford, 1991: 139-153.

Knight, R.T., Hillyard, S.A., Woods, D.L. and Neville, H.J. The effects of frontal cortex lesions on event-related potentials during auditory selective attention. Electroenceph. clin. Neurophysiol., 1981, 52: 571-582.

Knight, R.T., Scabini, D. and Woods, D.L. Prefrontal cortex gating of auditory transmission in humans. Brain Res., 1989, 504: 338- 342.

Kwon, S.E. and Heilman, K.M. Ipsilateral neglect in a patient following a unilateral frontal lesion. Neurology, 1991, 41: 2001- 2004.

Lang, A.H., Nyrke, T., Ek, M., Aaltonen, O., Raimo, I. and N~i~it~inen, R. Pitch discrimination performance and auditive event-related potentials. In: C. Brunia, A. Gaillard and A. Kok (Eds.), Psy- chophysiological Brain Research, Vol. 2. Tilburg University Press, Tilburg, 1990: 294-298.

Lounasmaa, O.V., Hari, R., Joutsiniemi, S. and H~im~il~iinen, M. Multi-squid recordings of human cerebral magnetic fields may give information about memory processes. Europhys. Lett., 1989, 9: 603-608.

Lyytinen, H., Blomberg, A.P. and N~i~it~inen, R. Event-related potentials and autonomic responses to a change in unattended auditory stimuli. Psychophysiology, 1992, 29: 523-534.

N~i~it~inen, R. Orienting and evoked potentials. In: H. Kimmel, E. Van Olst and J. Orlebeke (Eds.), The Orienting Reflex in Hu- mans. Erlbaum, Hillsdale, NJ, 1979: 61-75.

N~i~it~inen, 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.

362 K. ALt tO El AL.

N~i~it~inen, R. Mismatch negativity outside the strong attentional focus: a commentary on Woldorff et al, (1991). Psychophysiology, 1991, 28: 478-484.

N~i~it~inen, R. Attention and Brain Function. Erlbaum, Hillsdale, N J, 1992.

N~i~itiinen, R. and Pictom T.W. The NI 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~ifit~inen, R., Galliard, A,W.K. and M~intysalo, S. Early selective attention effects on the evoked potential reinterpreted. Acta Psychol. (Amst.), 1978, 42: 313-329.

N~i~it~inen, R., Simpson. M. and Loveless, N.E. Stimulus deviance and evoked potentials. Biol. Psychol., 1982, 14: 53-98.

N~i~it~inen, R., Paavilainen, P., AIho, K., Reinikainen, K. and Sams, M. Do event-related potentials reveal the mechanism of thc auditory sensory memory in the human brain'? Neurosci. l,ett.. 1989, 98: 217-221.

Novak, G.P., Ritter, W., Vaughan, H.G, and Wiznitzer, M.L. Differ- entiation of negative event-related potentials in an auditory discrimination task. Electroenceph. din. Neurophysiol., 1990, 75: 255-275.

Novak, G., Ritter, W. and Vaughan, H.G. Mismatch detection and the latency of temporal judgements. Psychophysiology, 1992, 2tk 398-411.

Paavilainen, P., Alho, K., Reinikainen, K., Sams, M. and N~i~itiinen. R. Right hemisphere dominance of different mismatch negativi- ties. Electroenceph. clin. Neurophysioi., 1991, 78: 466-479.

Sams, M., H~im/il~iinen, M., Antervo, A., Kaukoranta, E., Reini- kainen. K. and Hari, R. Cerebral neuromagnetic responses ew,ked by short auditory stimuli. Electroenceph. clin. Neurophysiol.. t985a, 61: 254-266.

Sams, M., Paavilainen, P,, Alho, K. and N~i~it~inen, R. Auditory frequency discrimination and event-related potentials. Electroen- ceph. clin. Neurophysiol., 1985b, 62: 437-448.

Sams, M., Kaukoranta. E., ttS.miil~inen, M. and N:,iiit~inen, R. Corti- cal activity elicited by changes in auditory stimuli: different sources for the magnetic Nl00m and mismatch responses. Psy- chophysiology, 1991, 28: 21-29.

Scabini, D., Knight, R.T. and Woods, D.L. Frontal lobe conlribu+ lions to the human P3a, Soc. Neurosci. Abst., 1989, 15: 477.

Scherg, M., Vajsar, J. and Picton, T. A source analysis of human auditor3 evoked potentials. J. Cogn. Neurosci., 1989, 1: 336-355,

Selemon, L.D. and Goldman-Rakic, P.S. Common cortical and sub- cortical targets of the dorsolateral prefrontal and posterior parietal cortices in the rhesus monkey: evidence for a distributed

neural network subserving spatially guided behavior. J. Neurosci., 1988, 8: 4049-4068.

Singh, J, and Knight, R,rl ', Frontal lobe contribution to voluntary movements in humans. Brain Res., 19911, 531: 45-54.

Skinner, J.E. and Yingling, C.D. Central gating mechanisms that regulate event-related potentials and behavior. Prog. Clin. Neu- rophysiol., 1977, 1 :30-69

Sokolov, E.N. The neuronal mechanisms of the orienting reflex. In: E. Sokokw and O. Vinogradova (Eds.), Neuronal Mechanisms of lhe Orienting Reflex. Erlbaum, tIillsdale, N J, 1975: 217-235.

Tiitinen, H., Alho, K., Huotilainen, M., Ilmoniemi, R.J.. Simota, J. and NZ, i~it~inen, R. Tonotopic auditory cortex and the magnetoen- cephalographic (MEG) equivalent of the mismatch negativity, Psychophysiology, 1993, 30: 537-54(I.

Wilson, F.A., O'Scalaidhe, S.P. and Goldman-Rakic, P.S. Dissocia- tion of object and spatial processing domains in the primate prefrontal cortex. Science. 1!193, 260: 1955-.1958.

Woldorff, M.G. Distortion of ERP averages due to overlap from temporally adjacent ERPs: analysis and correction. Psychophysi- ology, 1993, 30: 98--119.

Woldorff, M.G., ttacklcy, S.A. and Itillyard, S.A. The ¢flects ol channel-selective attention on the mismatch negativity wave elicited by deviant tones. Psychophysiology, 1991.28: 3{/-42,

Woods, D.L. The component structure of the N I wave of lhe human auditory evoked potential. Electroenceph. clin. Neurophysiol., t994, Suppl., in press.

Woods, D.I,. and Clayworth, C.C. Click spatial position influences middle latency auditory evoked potentials (MAEPs) in humans, Electroenceph. clin. Neurophysiol.. 1985, 60:122 129.

Woods, D.L. and Knight. R.T Eleclrophysiologic evidence of iw creased distraclibilit5 after dorsolateral prefrontal lesions. Neu- roh)gy, 1986, 36:212 21~,

Woods. D.L., Alho, K. and Algazi, A. lntermodal selective attention 1. Effects on event-related potentials to lateralized auditor}, and visual stimuli. Electroenceph. clin. Neurophysiol., 1992, 82: 341- ~55.

Woods, D,L.. Knight, [4,, 1' arid Scabini, 1). All:.ttonsictd substratc~ 151 auditory selective atlentlon: behavioral and electrophysmlogical effects o1 posterior association cortex lesions, ('ogn. Brain Res., 1993. I: 227-24(t.

Yamaguchi, S. and Knight. R.T, (iating of somatosensor3 inputs b,. human prefrontal cortex. Brain Res , 1990, 52l: 281 288,

Yetcrian, E.H. and Pan@a. I). Corticothalamic connections (51 1he posterior parietal cortex in the rhesus monkey. J. Comp. Neurol., 1985. 237: 408-426.

Lesions of Frontal Cortex Diminish the Auditory Mismatch Negativity

Documents

Transcript of Lesions of Frontal Cortex Diminish the Auditory Mismatch Negativity