Journal of Clinical Neurophysiology 20(5):351–360, Lippincott Williams & Wilkins, Inc., Philadelphia © 2003 American Clinical Neurophysiology Society

Event-Related Potentials in the Frontal Lobe During Performance of a Visual Duration Discrimination Task Isabelle Paul, Christophe Le Dantec, Christian Bernard, Robert Lalonde, and Mohamed Rebaï Université de Rouen, Faculté des Sciences, Laboratoire Psychologie et Neurosciences de la Cognition, Rouen Cedex, France

Summary: An event-related potential (ERP) study was conducted to elucidate the role of the prefrontal cortex (PFC) in time estimation. Subjects discriminated between three pairs of visual stimuli lasting from 100 ms and 2 seconds by determining whether the second stimulus was longer or briefer than the first. Event-related potentials were recorded in frontal and prefrontal regions after offset of the second stimulus (S2). The results indicated that the accuracy of the performances depended on stimulus duration and presentation order. In the brief-long order, the number of successful responses was higher as a function of stimulus duration. A time-related late positive component (LPCt) was revealed at prefrontal and frontal electrodes whose latency and amplitude differed depending on stimulus duration and order. The amplitude of this positive wave was higher when performance levels increased in the brief-long but not the reverse order. These results indicate that the LPCt may reflect successful decision-making or retrieval during time estimation as a result of neuronal activity in the PFC. Key Words: Time estimation—Duration discrimination—Visual discrimination—Prefrontal cortex.

although the susceptibility of the comparator to reinforcement was described by Wearden and Grindod (2003), adding to the importance of memory (Monfort et al., 2000) and selective attention (Brown and Boltz, 2002; Burle and Casini, 2001; Macar et al., 1994; Nobre, 2001) in temporal information processing. Based on the Church model (1984), several brain regions have been implicated in time estimation, including the neocortex, the hippocampus, the basal ganglia, the thalamus, and the cerebellum (Harrington et al., 1998a,b; Ivry and Keele, 1989; Lalonde and Hannequin, 1999). It is highly probable that these brain regions are involved in different aspects of temporal information processing, but the precise role of each has not yet been determined. For example, the cerebellum may be responsible for the pacemaker function, while the prefrontal cortex (PFC) may be involved in working memory or selective attention (Mangels et al., 1998). In support of this hypothesis is the finding that patients with cerebellar lesions were impaired in the discrimination of interval durations between sounds irre-

According to the Church (1984) model, three levels appear necessary for successful time estimation. The first level concerns the internal clock, represented as a pacemaker that produces impulses. The second level holds in memory the impulses counted in an accumulator. In the final stage, a comparator is involved in a decision process. The neurobiological basis of these mental processes has been initiated with pharmacological studies of substances acting on the speed of the internal clock (Meck, 1996; Odum, 2002). The underlying basis of the memory component has also been undertaken (Monfort et al., 2000). According to Church (1984), memory for durations can be subdivided into working and reference subcomponents. However, few studies have focused on the comparator, without which no decision is possible,

Address correspondence and reprint requests to Dr. Mohamed Rebaï, Université de Rouen, Faculté des Sciences, Laboratoire Pshchologie et Neurosciences de la Cognition (EA1780), 76821 Mont-Saint-Aignan, Rouen Cedex, France; e-mail: [email protected].

351

352

I. PAUL ET AL.

spective of their length, whereas patients with PFC lesions were impaired only for the longest interval (Mangels et al., 1998). It may be assumed that a dysfunctional central pacemaker disrupts any estimate of time, while working memory and attention are more strongly engaged with increasing durations. Electrophysiological studies in monkeys have revealed that neuronal activity in the dorsolateral PFC and the anterior cingulate increased during the interval between the presentation of two stimuli and diminished before responding, concordant with a pacemaker function (Niki and Watanabe, 1979). In addition to these regions, the basal ganglia have been proposed as a possible pacemaker (Ivry, 1996). The role of ascending dopaminergic fibers in temporal information processing has been underscored by findings in normal subjects receiving neuroleptic drugs (Rammsayer, 1997; Rammsayer and Vogel, 1992) as well as patients with Parkinson’s disease (Artieda et al., 1992; Harrington et al., 1998a; Pastor et al., 1992), schizophrenia (Densen, 1977; Wahl and Sieg, 1980), and attention deficit disorder (Smith et al., 2002). Event-related potentials (ERP) studies have contributed to the possible implication of the frontal lobe in the memory component of temporal information processing. Schubotz and Friederici (1997) recorded ERPs while normal subjects retained either temporal or spatial information. The topographic patterns of brain activity differed depending on the task used; a slow negative wave was prominent in occipitoparietal regions for the spatial task and a different slow negative wave was prominent in frontal regions for the temporal task. Casini and Macar (1996, 1999), Monfort et al. (2000), and Pouthas et al. (2000) have identified during time estimation the well-known contingent negative variation (CNV) originally described by Walter et al. (1964). The CNV is observable not only during temporal processing but also in many other experimental conditions, provided recordings are made during a fixed interval between the presentation of two stimuli. It is therefore generally regarded as an electrophysiologic correlate of expectancy. According to Macar (1977), the CNV is involved in the temporal processing underlying the intention of initiating a motor response at a specific period. Moreover, the higher CNV amplitude seen during time estimation on incorrect than on correct trials (Casini and Macar, 1999; Ladanyi and Dubrovsky, 1985) indicates that accurate time estimation may be dependent on a minimal recruitment of prefrontal cortex (PFC) neurons, underscoring the role of this component in the attentional processes of the accumulator. Based on these findings, the PFC may be a key structure involved in memory or selective attention and its activity seems to be associated with behavioral performances. The first goal of the present study was to identify frontal lobe-related ERPs recorded during a J Clin Neurophysiol, Vol. 20, No. 5, 2003

visual duration discrimination task. The second goal was to determine whether these ERPs are associated with behavioral performances. We focused on the decision process in the comparator, the third level of temporal information processing according to the Church (1984) model. The subjects were presented with one of three stimulus duration pairs (100/200 ms; 300/600 ms, and 1000/2000 ms) in either order and were asked to judge whether the second stimulus was longer or shorter than the first. The accuracy of the responses was dependent on stimulus duration and order. We therefore analyzed whether the ERPs were related to task performance by determining in which condition the amplitude of the ERPs was maximal. By contrast to previous studies in which ERPs were measured either during S2 or the S1-S2 interval (Casini and Macar, 1999; Ladanyi and Dubrovsky, 1985; Monfort et al., 2000; Pouthas et al., 2000), the ERPs in the present investigation were measured after S2 offset. A part of the present report was presented in abstract form (Paul et al., 2002). METHODS Subjects Twenty-nine subjects (24 women and 5 men) volunteered for the behavioral part of the study, among whom 10 (5 men and 5 women) volunteered as well for the EEG part, which was conducted in a separate session. The majority of subjects consisted of university students, with a mean age of 23 years, normal or corrected-to-normal eyesight, and they were right-handed according to criteria established by Oldfield (1971). The research protocol followed the guidelines of the European Council Directive (86/609/EEC) for the ethical treatment of human subjects. Temporal Discrimination Task In the paradigm without EEG, the subjects saw one of three stimulus duration pairs (100/200 ms; 300/600 ms, and 1000/2000 ms) in either order on an IBM-compatible computer screen and were asked to judge whether the second stimulus (S2) was longer or shorter than the first (S1) by pressing on a keyboard key marked with an arrow pointing either right (3) or left (4). Whenever S2 was longer, the (3) key had to be selected with the right middle finger, and whenever S2 was shorter, the (4) key had to be selected with the right index finger. The stimuli consisted of a 1 pixel-sized white spot set on a dark background with a contrast sensitivity of 99%. The interstimulus (S1/S2) interval was held constant at 1500 ms. Each of the three stimulus pairs was presented 20 times in a pseudorandom

PFC IN TIME ESTIMATION order. The number of correct responses and the reaction times (RTs) for correct responses were recorded. In the paradigm with EEG, the same procedure was followed, except that a beep sound was presented 1 second after S2 offset to exclude response-related ERPs (Rockstroh et al., 1982). Each of the stimulus pairs was presented a higher number of times (100) to permit valid statistical analyses of ERPs, and no RTs were taken because of the need for including the auditory beep condition. Event-Related Potentials (ERPs) The EEG was recorded with a 32-electrode array displayed according to the 10 to 20 international system, with the fronto-central electrode as the reference for data accumulation. Event-related potentials for each electrode were then determined as a function of a mean reference calculated from the following 20 electrodes: F3, C3, P3, F7, T3, T5, CP3, TP7, F4, C4, P4, F8, T4, T6, CP4, TP8, FZ, PZ, CZ, and CPZ (Bertrand et al., 1985). The signals were amplified, digitalized, sampled (1 point per 3.92 ms), filtered (0.1 Hz to 100 Hz), and stored on an IBM-compatible computer with Deltamed (Paris, France) software. The baseline was calculated on the basis of the average amplitude of the 250 ms period preceding S2 onset. Impedance was set at 5 K⍀. The EEG was continuously recorded during the experiment, and codes, synchronized to stimulus delivery, were used to average epochs off-line. All epochs correspond to 3 seconds after stimulus onset (the longest duration was 2 seconds plus the 1-second interval between S2 offset and the auditory tone). During wave averaging, artifacts from ocular movements (criterion⬎100 ␮V) were eliminated. After grand averaging, the data were filtered at 48 Hz and visualized in the form of electrophysiologic signals or in the form of topographic mapping. As illustrated in Fig. 1, the temporal windows were determined as a function of latencies and duration of the positive wave

353

appearing after S2 offset. Since these two characteristics depended on stimulus duration and order (see below), a specific temporal window was established for each experimental condition. The mean amplitude was calculated within the temporal windows (1 point per 3.92 ms as a function of the baseline). Afterward, temporal windows were selected for each stimulus pair. For the 100/200 and 200/100 ms discriminations, the optimal window was 600 to 1000 ms and 700 to 1150 ms, respectively. For the 300/600 and 600/300 ms discriminations, the optimal window was 900 to 1150 ms and 750 to 1000 ms, respectively. For the 1/2 second and 2/1 second discriminations, the optimal window was 2200 to 2500 ms and 1350 and to 1550 ms, respectively. Statistical Analyses The behavioral data were analyzed by 3 (stimulus duration) ⫻ 2 (stimulus order) within-subject analyses of variance (ANOVAs) for correct responses and RTs. Because two subjects did not have a single correct response for the 100/200 second discrimination, their RT data were excluded. Event-related potentials were analyzed by 3 (stimulus duration) ⫻ 2 (stimulus order) (ANOVAs) with the Greenhouse-Geisser correction applied in cases with 2 df or more. The seven electrodes included FP1 and FP2 in the prefrontal area, F3, F4, F7, and F8 in the lateral frontal area, and FZ in the fronto-central area, and each was analyzed separately. RESULTS Behavioral Results in the Paradigm Without EEG The stimulus duration ⫻ presentation order interaction was significant for correct responses (F[2,56]⫽75.31; P ⬍ 0.001) and RTs (F [2,52]⫽50.3; P ⬍ 0.001). As

FIG. 1. The 1/2 second condition used as an example of the criteria used for the determination of temporal windows. The LPCt appears after S2 offset. The choice of the temporal window depends on the onset and duration of this component.

J Clin Neurophysiol, Vol. 20, No. 5, 2003

354

I. PAUL ET AL.

FIG. 3. Mean ⫾ SEM number of correct responses in subjects performing the visual duration discrimination task in the paradigm with EEG for brief (100/200 or 200/100 ms), intermediate (300/600 or 600/300 ms), and long (1/2 second or 2/1 second) stimulus duration pairs.

tion (F[1,26]⫽127.1; P ⬍ 0.001), while no effect was seen for the intermediate duration. Behavioral Results in the Paradigm With EEG

FIG. 2. Mean ⫾ SEM number of correct responses (A) and reaction times (B) in subjects performing the visual duration discrimination task in the paradigm without EEG for brief (100/200 ms or 200/100 ms), intermediate (300/600 or 600/300 ms), and long (1/2 second or 2/1 second) stimulus duration pairs.

shown in Fig. 2A, the number of correct responses was higher for the 200/100 ms discrimination than for the 100/200 ms discrimination (F[1,28]⫽102.7; P ⬍ 0.001). This effect was reversed for the longest stimulus duration, as the 1/2 second subtask was easier to discriminate than 2/1 second (F[1,28]⫽49.8; P ⬍ 0.001). In contrast, the success rate of the intermediate duration was not affected by the order of presentation (300/600 second or 600/300 second). In the brief-long order, the curve relating percentage correct increased as a function of increasing durations, whereas for the long-brief order, the curve was flat. As shown in Fig. 2B, the RT results were identical, as shorter RTs were associated with a higher number of correct responses. The subjects had lower RTs for the 200/100 ms discrimination than for the 100/200 ms (F[1,28]⫽13.7; P ⬍ 0.001) discrimination and lower RTs for the 1/2 second discrimination than 2/1 for the 2/1-second discriminaJ Clin Neurophysiol, Vol. 20, No. 5, 2003

As seen in Fig. 3, the statistical results for correct responses by the smaller group of subjects undergoing EEG evaluation did not differ from those of the larger group not undergoing EEG evaluation, as the duration ⫻ order interaction was still significant (F[2,18]⫽32.14; P ⬍ 0.001). As mentioned above, the subjects undergoing EEG testing could not be evaluated for RTs. EEG Analyses After S2 offset, a time-related late positive component (LPCt) appeared at prefrontal (FP1 and FP2) and frontal (F3, F4, F7, F8, FZ) electrodes (Fig. 4). The amplitude of the LPCt was much greater in the prefrontal region than in the frontal region. Latencies before appearance of the wave differed depending on stimulus duration and order, and therefore separate time windows were used (Fig. 5). Irrespective of wave latencies, the LPCt ended a few dozen ms before the mean RT value measured in the larger group of subjects. The LPCt amplitudes illustrated in Figs. 6, 7, and 8 refer to the values based on the mean of each temporal window. The amplitudes indicated in Figs. 6, 7, and 8 represent the calculated mean relative to baseline from points sampled every 3.92 ms within the temporal window. Thus, the values of the mean amplitudes obtained in this fashion were relatively weaker than the ones measured between the baseline and the LPCt peak.

PFC IN TIME ESTIMATION

355

FIG. 4. Electrophysiologic activity recorded at FP1, FP2, F3, F4, F7, F8, and FZ electrodes in the brief-long order of the visual duration discrimination task for long durations.

Analyses of the amplitude of the LPCt were undertaken for the different stimulus durations and presentation order for the seven electrodes (Table 1). The stimulus duration ⫻ presentation order interaction was significant for every electrode except F8, as LPCt amplitudes increased for longer stimulus durations in the brief-long order (100/200 ms, 300/600 ms, 1/2 second), but not in the reverse order. Figure 6 illustrates the amplitudes based on the mean of each temporal window for the two prefrontal and two lateral frontal electrodes. These greater amplitudes coincided with a higher number of correct

responses and lower RTs (Fig. 7), but once again, this was true only for the brief-long (100/200 ms, 300/600 ms, 1/2 second), and not the reverse order. There was no interhemispheric difference for any of the analyses.

Brain-Behavior Relations As revealed by one-way ANOVA (Table 2 and Fig. 8), LPCt amplitudes were associated not only with longer stimulus durations but also with a higher percentage of

FIG. 5. Prefrontal (FP1 and FP2) eventrelated potentials for brief (A), intermediate (B), and long (C) durations for the brief-long (bold lines) and long-brief (normal lines) presentation order. The characteristics of the LPCt change as a function of stimulus duration and presentation order.

J Clin Neurophysiol, Vol. 20, No. 5, 2003

356

I. PAUL ET AL.

FIG. 6. Amplitude variations of prefrontal (FP1, FP2) and lateral frontal (F3, F4) electrodes as a function of stimulus duration and order.

correct responses. This relationship was found only in the brief-long and not the reverse order. DISCUSSION Duration Discrimination Task The success rate of subjects performing temporal discriminations depended on the stimulus duration and the order of presentation. The number of correct responses was higher for the 200/100 ms discrimination than for the 100/200 ms discrimination. This effect was reversed for the longest stimulus duration, as the 1/2 second condition was easier to discriminate than 2/1 second. In contrast, the success rate of the intermediate duration

(300/600 ms or 600/300 ms) was not affected by the presentation order. The RT results mirrored those found for number of correct responses. For the brief-long order, performance levels were higher as a function of increasing stimulus durations, whereas for the long-brief order, the curve was flat. Two possible reasons may be offered for explaining the superior level achieved for the 1/2 second subtask over 2/1 second. It is possible that the relative difficulty of the 2/1 second discrimination was caused by time-related memory interference. At these long durations, it is perhaps easier to retain over time a briefer stimulus than a longer one. Since S1 was longer than S2 for the 2/1 second discrimination, the S1 duration may have been more difficult to retain than the

FIG. 7. Elevated ERP amplitudes as a function of higher correct responses and lower reaction times at the FP1 electrode in the brief-long order. ERP, event-related potential.

J Clin Neurophysiol, Vol. 20, No. 5, 2003

PFC IN TIME ESTIMATION TABLE 1. Two-way ANOVA results of wave amplitude for each electrode site depending on presentation order (O), stimulus duration (D), or the interaction between the two Factor (d.f.) FP1 O (1,9) D (2,18) OxD (2,18) FP2 O (1,9) D (2,18) OxD (2,18) F3 O (1,9) D (2,18) OxD (2,18) F4 O (1,9) D (2,18) OxD (2,18) F7 O (1,9) D (2,18) OxD (2,18) F8 O (1,9) D (2,18) OxD (2,18) FZ O (1,9) D (2,18) OxD (2,18)

F

Epsilon

357

TABLE 2. Trend analyses relating LPCt amplitudes and either stimulus duration or behavioral performances at seven electrode sites for the brief-long (B-L) and the long-brief (L-B) presentation order

P Correct responses

13.22 5.67 19.59

— 1.00 0.87

0.0054 0.0123 0.0001

24.38 5.62 13.5

— 0.92 0.83

0.0008 0.0153 0.0007

15.86 6.56 4.34

— 0.87 0.88

0.0032 0.0105 0.0357

F3

9.11 4.54 7.12

— 0.83 0.88

0.0145 0.0341 0.0078

F7

13.92 0.56 4.21

— 0.88 0.79

0.0047 0.5588 0.0441

FZ

14.47 3.47 1

— 0.86 0.68

0.0042 0.0624 0.3658

12.36 2.38 5.73

— 0.84 0.98

0.0066 0.1326 0.0124

ANOVA, analysis of variance.

reverse condition, as S1 is farther away in time from the retention test than S2. This may not be true for briefer durations. The second possible explanation is that the 1-second duration may have appeared shorter when presented as S1 than S2 due to a time-compression bias. Subjective shortening as a function of the S1-S2 interval was described by Wearden et al. (1993, 1999), in which subjects judged S1 to be of a shorter duration than S2 with increasing S1-S2 intervals. Although the S1-S2 interval in the present study remained fixed, it is possible that the 1 second duration appeared shorter as S1 than S2 because S1 was farther away in time relative to the retention test. If the second explanation is correct, a similar effect should be discerned at stimulus durations longer than 2 seconds, a hypothesis we intend to evaluate in the future. A different hypothesis must be formulated for explaining the relative ease of the 200/100 ms discrimination over 100/200 ms. For these brief durations at the limit of the subjects’ capacities, the 200/100 ms discrimination may have appeared easier because of a response bias in reporting S2 as shorter.

Electrodes FP1 FP2

F4

F8

Order

F (2,18)

Epsilon

B-L L-B Order B-L L-B B-L L-B B-L L-B B-L L-B B-L L-B B-L L-B B-L L-B

91.46 0.62 F(2,18) 15.33 0.01 13.55 0.00 11.80 0.42 6.20 0.68 4.76 0.26 8.76 0.19 8.76 0.19

0.93 0.00 Epsilon 0.84 0.85 0.91 0.71 0.97 0.90 0.81 0.89 0.94 0.87 0.68 0.67 0.89 0.95

P 1.5E-09 0.4914 P 0.0004 0.9882 0.0004 0.9837 0.0006 0.6444 0.0019 0.7173 0.0104 0.5008 0.0402 0.6886 0.0034 0.8167

LPCt, time-related late positive component.

Event-Related Potentials (ERPs) and Behavioral Performances During performance of the visual duration discrimination task, a time-related late positive component (LPCt) appeared after S2 offset. The LPCt amplitude was higher at prefrontal sites than at frontal electrode sites, concordant with the possible existence of a neuronal generator in the PFC as opposed to motor-related regions. The amplitude and latency differed depending on stimulus duration and presentation order. The stimulus duration ⫻ presentation order interaction was significant for six out of seven electrodes: namely FP1, FP2, F3, F4, F7, and FZ. The LPCt amplitude increased as a function of stimulus duration, but this effect was only seen for the brief-long (100/200 ms, 300/600 ms, 1 second/2 second) and not the long-brief (200/100 ms, 600/300 ms, 2 second/1 second) presentation order. As mentioned above, it was only in the brief-long order that longer stimulus durations resulted in improved behavioral performances. The greater wave amplitudes coincided with a higher success rate and lower RTs. In the long-brief order, the curve relating success rate and stimulus duration was flat, and there was no apparent relation between ERP amplitudes and behavioral performances. These results are concordant with the hypothesis that the PFC is involved in duration discriminations, as proposed by Fuster (1995) and Gibbon et al. (1997). Other investigators have previously reported frontal lobe-related ERPs, but the polarity of those waves was negative, J Clin Neurophysiol, Vol. 20, No. 5, 2003

358

I. PAUL ET AL.

FIG. 8. Relations between ERP amplitudes and the number of correct responses at FP1, FP2, F3, and F4 electrodes. The amplitudes and the number of correct responses increase with stimulus duration, especially at prefrontal sites, only for the brief-long order (A) and not the long-brief (B) order. ERP, event-related potential.

a result that may be explained by the type or timing of those measurements. Schubotz and Friederici (1997) described a negative wave in frontal regions during performance of a time estimation task. The selectivity of this component for time-related processes was demonstrated by the finding that the cartographic pattern of the temporal task was distinct from that of a spatial task. In a similar manner, Ladanyi and Dubrovsky (1985), Casini and Macar (1999), Monfort et al. (2000), and Pouthas et al. (2000) described a late negative wave during performance of a time estimation task. As with our late positive wave, the CNV was related to task performance, except that higher amplitudes were related to poor and not to better performances. In particular, the amplitude of the CNV was smaller during correct as opposed to incorrect trials (Casini and Macar, 1999; Ladanyi and Dubrovsky, J Clin Neurophysiol, Vol. 20, No. 5, 2003

1985), indicating that this wave may be an error signal. The lower amplitudes seen during successful trials may indicate a more efficient recruitment of neurons in the PFC. The positive polarity of our time-related component may be explained by the period at which the ERPs was measured. By contrast to ERPs measured either during S2 or the S1-S2 interval (Casini and Macar, 1996, 1999; Ladanyi and Dubrovsky, 1985; Monfort et al., 2000; Pouthas et al. 2000), the ERPs in the present investigation were measured after S2 offset. The CNV may reflect attentional or encoding processes. In addition, this negative wave may reflect retrieval, because it appeared during presentation of a stimulus duration while subjects were comparing it to remembered standard durations (Monfort et al. 2000). The LPCt may also reflect retrieval or else those decisional processes directly related to S1-S2 discriminations. By contrast to its possible role as an error signal in temporal tasks, the CNV amplitude was associated with superior performances in nontemporal tasks. In a twochoice RT task, faster subjects had higher CNVs (Hohnsbein et al. 1998). The amplitude of the CNV is also dependent on task requirements. In a go/no go task, the CNV amplitude was higher during the go as opposed to the no go part, indicating its sensitivity to the nature of the response pattern (Davies et al. 2001). These results show that changes in CNV amplitudes are not specific to temporal tasks. It remains to be determined whether that is the case for the LPCt. If so, this wave may constitute a signal for the comparator of the internal clock (Church, 1984). The LPCt was not associated with interhemispheric differences. This result is in line with those of regional brain metabolism measured by functional magnetic resonance imaging, indicating PFC-related changes during duration discriminations (Basso et al., 2003). However, it remains to be determined to what extent these findings are generalizable. On one hand, a preferential righthemisphere involvement was reported during time estimation, as measured by regional brain metabolism (Lejeune et al., 1997; Maquet et al., 1996; Pouthas et al., 2000; Rao et al., 2001) and ERPs (Monfort et al., 2000; Pouthas et al., 2000). Moreover, patients with right-sided lesions were more impaired in temporal tasks than those with left-sided lesions (Harrington et al., 1998b; Kagerer et al., 2002). On the other hand, a bilateral or even a preferential left-sided involvement during temporal tasks was reported in several brain regions, as determined by regional brain metabolism (Coull and Nobre, 1998; Kawashima et al., 2000; Rubia et al., 1998). It appears likely that the apparent differences are caused by task-

PFC IN TIME ESTIMATION related factors and that both hemispheres are involved in different aspects of time estimation. Timsit-Berthier and Gérono (1998) postulated that the CNV reflects neuronal assemblies within a network, whereas late positive waves reflect disassemblies. It may be hypothesized that the LPCt is involved in decision processes by filtering out irrelevant information. The LPCt would then be the result of increased neuronal inhibition of networks corresponding to incorrect choices. The finding of superior performances with higher LPCt amplitudes and longer stimulus durations, at least for the brief-long presentation order, would thus be expected. The absence of such a pattern in the reverse presentation order may underlie the lower performance seen at the longest stimulus duration. A different information processing system may be at work for shorter durations, reflecting perceptual as opposed to cognitive operations (Rammsayer and Grondin, 2000). CONCLUSIONS These results show that the LPCt, appearing during decision making, may be an electrophysiological correlate of temporal information processing. The amplitude and latency of this component depends on stimulus duration and presentation order. These modulations may be due to the different cognitive load underlying these subtasks. Our results also indicate that LPCt amplitudes are probably linked to behavior and may therefore be used as an index of optimal performance. Additional studies are necessary to discover to what extent these results are generalizable in other temporal and nontemporal tasks. REFERENCES Artieda J, Pastor MA, Lacruz F, Obeso JA. Temporal discrimination is abnormal in Parkinson’s disease. Brain 1992;115:199 —210. Basso G, Nichelli P, Wharton CM, Matthew P, Grafman J. Distributed neural systems for temporal production: a functional MRI study. Brain Res Bull 2003;59:405—11. Bertrand O, Perrin F, Pernier J. A theoretical justification of the average reference in topographic evoked potential studies. Electroencephalogr Clin Neurophysiol 1985;62:462— 4. Brown SW, Boltz MG. Attentional processes in time perception: effects of mental workload and event structure. J Exp Psychol Human Percept Perform 2002;28:600 —15. Burle B, Casini L. Dissociation between activation and attention effects in time estimation: implications for internal clock models. J Exp Psychol Human Percept Perform 2001;27:195—205. Casini L, Macar F. Can the level of prefrontal activity provide an index of performance in humans? Neurosci Lett 1996;219:71— 4. Casini L, Macar F. Multiple approaches to investigate the existence of an internal clock using attentional resources. Behav Proc 1999; 45:73— 85. Church RM. Properties of the internal clock. Ann N Y Acad Sci 1984;423:566 — 82.

359

Coull JT, Nobre AC. Where and when to pay attention: the neural systems for directing attention to spatial locations and to time intervals as revealed by both PET and fMRI. J Neurosci 1998;18: 7426 —35. Davies PL, Segalowitz SJ, Dywan J, Pailing PE. Error-negativity and positivity as they relate to other ERP indices of attentional control and stimulus processing. Biol Psychol 2001;56:191—206. Densen ME. Time perception and schizophrenia. Percept Motor Skills 1977;44:436 — 8. Fuster JM. Temporal processing. Ann N Y Acad Sci 1995;769:173— 81. Gibbon J, Malapani C, Dale CL, Gallistel CR. Toward a neurobiology of temporal cognition: advances and challenges. Curr Op Neurobiol 1997;7:170 — 84. Harrington DL, Haaland KY, Hermanowicz N. Temporal processing in the basal ganglia. Neuropsychology 1998a;12:3—12. Harrington DL, Haaland KY, Knight RT. Cortical networks underlying mechanisms of time perception. J Neurosci 1998b;18:1085—95. Hohnsbein J, Falkenstein M, Hoormann J. Performance differences in reaction tasks are reflected in event-related brain potentials (ERPs). Ergonomics 1998;41:622—33. Ivry R, Keele SW. Timing functions of cerebellum. J Cogn Neurosci 1989;1:136 —52. Ivry R. The representation of temporal information in perception and motor control. Curr Op Neurobiol 1996;6:851—7. Kagerer FA, Wittmann M, Szelag E, Steinbüchel N. Cortical involvement in temporal reproduction: evidence for differential roles of the hemispheres. Neuropsychologia 2002;40:357— 66. Kawashima R, Okuda J, Umetsu. A et al. Human cerebellum plays an important role in memory-timed finger movement: an fMRI study. J Neurophysiol 2000;83:1079 — 87. Ladanyi M, Dubrovsky B. CNV and time estimation. Int J Neurosci 1985;26:253—7. Lalonde R, Hannequin D. The neurobiological basis of time estimation and temporal order. Rev Neurosci 1999;10:151—73. Lejeune H, Maquet P, Bonnet M, Casini L, Ferrara A, Macar F, Pouthas V, Timsit-Berthier M, Vidal F. The basic pattern of activation in motor and sensory temporal tasks: positron emission tomography data. Neurosci Lett 1997;235:21— 4. Macar F. Signification des variations contingentes négatives dans la dimension temporelle du comportement. Ann Psychol 1977;77: 384 — 439. Macar F, Grondin S, Casini L. Controlled attention sharing influences time estimation. Mem Cognit 1994;22:673— 86. Mangels J, Ivry R, Shimizu N. Dissociable contribution to the prefrontal and neocerebellar cortex to time perception. Brain Res 1998; 7:15—39. Maquet P, Lejeune H, Pouthas V, et al. Brain activation induced by estimation of duration: a PET study. Neuroimage 1996;3:119 — 26. Meck WH. Neuropharmacology of timing and time perception. Brain Res 1996;3:227— 42. Monfort V, Pouthas V, Ragot R. Role of frontal cortex in memory duration: an event related potential study in humans. Neurosci Lett 2000;186:91— 4. Niki H, Watanabe M. Prefrontal and cingulate unit activity during timing behavior in the monkey. Brain Res 1979;171:213—24. Nobre AC. Orienting attention to instants in time. Neuropsychologia 2001;39:1317—28. Odum AL. Behavioral pharmacology and timing. Behav Processes 2002;57:107—20. Oldfield RC. The assessment of analysis of handeness: the Edinburg Inventory. Neuropsychologia 1971;9:97—113. Pastor MA, Artieda J, Jahanshahi M, Obeso JA. Time estimation and reproduction is abnormal in Parkinson’s disease. Brain 1992;115: 211—25. Paul I, Le Dantec C, Bernard C, Lalonde R, Rebaï M. Correlational analysis of event-related potentials in the prefrontal cortex and behavioral performance in a temporal discrimination task. FederJ Clin Neurophysiol, Vol. 20, No. 5, 2003

360

I. PAUL ET AL.

ation of the European Neuroscience Society Abstracts no 74.17, Paris: Federation of the European Neuroscience Society, 2002. Pouthas V, Garnero L, Ferrandez AM, Renault B. ERPs, PET analysis of time perception: Spatial and temporal brain mapping during visual discrimination tasks. Hum Brain Mapp 2000;10:49 — 60. Rammsayer T, Grondin S. Psychophysic of human timing. In: Miller R, ed. Time and the brain: conceptual advances in brain research. Amsterdam: Hardwood Academic, 2000;157— 67. Rammsayer T, Vogel WH. Pharmacologic properties of the internal clock underlying time perception in humans. Neuropsychobiology 1992;26:71— 80. Rammsayer T. Are there dissociable roles of the mesostriatal and mesolimbocortical dopamine systems on temporal information processing in humans? Neuropsychobiology 1997;35:1:36 — 45. Rao SM, Mayer AR, Harrington DL. The evolution of brain activation during temporal processing. Nat Neurosci 2001;4,3:317—23. Rockstroh B, Elbert T, Lutzenberger W, Birbaumer N. The effects of slow cortical potentials on response speed. Psychophysiology 1982;19:211—17. Rubia K, Overmeyer S, Taylor E, Brammer M, Williams S, Simmons A,

J Clin Neurophysiol, Vol. 20, No. 5, 2003

Andrew C, Bullmore E. Prefrontal involvement in “temporal bridging” and timing movement. Neuropsychologia 1998;36:1283—93. Schubotz R, Friederici AD. Electrophysiological correlates of temporal and spatial information processing. Neuroreport 1997;8:1981— 6. Smith A, Taylor E, Rogers JW, Newman S, Rubia K.J. Evidence for a pure time perception deficit in children with ADHD. J Child Psychol Psychiat 2002;43:529 — 42. Timsit-Berthier M, Gérono A. Manuel d’interprétation des potentiels évoqués endogènes (P300 et VCN). Sprimont: Mardaga, 1998. Wahl OF, Sieg D. Time estimation among schizophrenics. Percept Mot Skills 1980;50:535— 41. Walter WG, Cooper R, Aldridge VJ, McCallum WC, Winter AL. Contingent negative variation: an electrical sign of association and expectancy in the human brain. Nature 1964;203:380 — 4. Wearden JH. “Beyond the field we know...”: exploring the scalar timing theory. Behav Processes 1999;45:3—21. Wearden JH, Ferrara A. Subjective shortening in humans’ memory for stimulus duration. Q J Exp Psychol B 1993;46:163— 86. Wearden JH, Grindod R. Manipulating decision processes in the human scalar timing system. Behav Processes 2003;61:47—56.