Missoner, Et Al 2006 i4m44

DOI 10.1007/s00702-005-0443-9 J Neural Transm (2006) 113: 1477–1486

Frontal theta event-related synchronization: comparison of directed attention and working memory load effects P. Missonnier1;2, M.-P. Deiber2, G. Gold4 , P. Millet1 , M. Gex-Fabry Pun3 , L. Fazio-Costa4 , P. Giannakopoulos2;5 , and V. Iba´n˜ez1 1 Neuroimaging Unit, Service of Geriatric Psychiatry, and 3 Unit of Clinical Research, Department of Psychiatry, University Hospitals of Geneva, 1225 Geneva, and 4 Department of Geriatrics, University Hospitals of Geneva, 1226 Geneva, and 5 Service of Old Age Psychiatry, University Hospitals of Lausanne, 1008 Prilly, Switzerland 2

Received July 11, 2005; accepted December 8, 2005 Published online April 11, 2006; # Springer-Verlag 2006

Summary. Early studies showed that longterm encoding and retrieval of new information is associated with modulation of the theta rhythm. More recently, changes in theta power amplitude over frontal electrode sites were reported during working memory, yet their relative significance in regard to attentional and memory processes remains unclear. Event-related synchronisation responses in the 4–7.5 Hz theta EEG frequency band was studied in 12 normal subjects performing four different tasks: two working memory tasks in which load varied from one (1-back task) to two (2-back task) items, an oddball detection (attention) task and a ive fixation task. A phasic theta increase was observed following stimulus apparition on all electrode sites within each task, with longer culmination peak and maximal amplitude over frontal electrodes. Frontal theta eventrelated synchronization (ERS) was of higher amplitude in the 1-back, 2-back and detection tasks as compared to the ive fixation task. Additionally, the detection task elicited

a larger frontal and central theta ERS than the 2-back task. By analyzing theta ERS characteristics in various experimental conditions, the present study reveals that early phasic theta response over frontal regions primarily reflects the activation of neural networks involved in allocation of attention related to target stimuli rather than working memory processes. Keywords: Working memory, EEG, eventrelated synchronisation, theta oscillations, attention. Introduction Brain oscillations reflect the synchronization of neuronal cell assemblies that participate in local or long-distance network systems. Alteration of oscillatory brain temporal dynamics can occur in different EEG frequency bands as a result of functional modulation during perceptive or cognitive tasks. Among these task-related oscillations, the human theta activity has been implicated in memory

1478

P. Missonnier et al.

performance (Doppelmayr et al., 2000; Klimesch, 1999), after hippocampal theta rhythm was originally described in memory processing in the rodent (O’Keefe, 1993; Winson, 1978). However, task-related human theta activity is widely recorded over the scalp and cannot be univocally attributed to the hippocampal region (for a review see Kahana et al., 2001). Klimesch and colleagues reported an increase in theta power during encoding and retrieval of new information but not during semantic memory performance (Klimesch, 1999; Klimesch et al., 1997), suggesting that theta activity may not be involved in all types of long-term memory processes. Recently, the influence of short-term memory, mainly working memory, on scalp theta activity has received increasing interest. Using taskrelated power or time frequency analysis, several authors have reported an increase of frontal theta amplitude as a function of working memory load (Gevins et al., 1997; Jensen and Tesche, 2002; McEvoy et al., 2001). Krause et al. (2000) specifically examined the temporal dynamics of theta activity during classical working memory n-back tasks by applying the event-related synchronisation (ERS) and desynchronisation (ERD) analysis procedure (Pfurtscheller, 1992, 2001). They observed that theta ERS amplitude was larger over frontal electrodes sites during the memory condition with the highest memory component (2-back task) and suggested that frontal theta synchronization could be an index of working memory load. Working memory is by definition closely related to attentional processes. Indeed, focused attention to incoming stimulus is mandatory to ensure correct encoding, a fundamental step for further successful working memory processes. Thus, the influence of attention on theta modulation in working memory cannot be ignored. There is evidence for strong evoked theta activity during neuropsychological paradigms involving focused attention (Basar-Eroglu et al., 1992). Further-

more, results from animal studies suggest that theta oscillations recorded in motor and prefrontal cortex could participate to the attention system (Demiralp et al., 1994). The aim of this study was to determine the respective contribution of working memory and attentional processes to theta activity. We used the ERD=ERS procedure (Pfurtscheller and Aranibar, 1977) to analyze the EEG power in the theta frequency band during two visual n-back working memory tasks and an oddball detection task. In addition, a ive fixation task was used to determine whether theta activity could be elicited in absence of any specific task demand. Subjects and methods Participants Twelve healthy, cognitively intact right-handed volunteers (aged 22–31, 26.75 2.86; 5 women, 7 men) participated in this study. All participants had normal or corrected-to-normal visual acuity, and none reported a history of head injury, neurological or psychiatric disorders. All participants were medication free and none exhibited alcohol or drug abuse. Informed consent was obtained from all subjects. The study was approved by the Ethical Committee of the University Hospitals of Geneva, and was in line with the Helsinki Declaration.

Experimental design The subjects, comfortably seated, watched a computercontrolled display screen at a distance of 57 cm. They viewed pseudo-random sequences of consonant and vowels common to the French alphabet, and pressed a computer-controlled button with their right index finger as soon as a target appeared (response trials). For non-target stimuli, no motor response was required (noresponse trials). In relation to the task, targets were defined either according to the oddball (rare event) or to the n-back design (Fig. 1). Stimuli consisted of white letters, Arial font (2 2.5 visual angle), with 10% grey noise, embedded in a 50% random noise grey rectangular background patch (6 6.7 visual angle). They were presented in the centre of the screen for 0.5 s, separated by 5 s intervals (onset to onset) during which a dot helped subjects maintain fixation.

Frontal theta event-related synchronization

1479

task. Reaction time and performance were systematically recorded, but no on performance was provided.

Electrophysiological recordings Continuous EEG (Micromed, Brain Quick system 98, Treviso, Italy) was recorded using twenty surface electrodes placed over the scalp according to the 10–20 international system, with linked earlobes as reference. Skin impedance was kept below 5 k. Electrophysiological signals were sampled at 1024 Hz, with a lower cut-off of 0.3 Hz and a upper cut-off of 100 Hz (DC amplifiers Micromed). The electrooculogram (EOG) was recorded using two pairs of bipolar electrodes in both vertical and horizontal directions. Single pulses (TTL) synchronized with stimulus onset were recorded and used off-line to segment the continuous EEG data into epochs time-locked to stimulus onset.

Data processing Fig. 1. Schematic representation of the four tasks. In the ive fixation task, letter series identical to the 2-back working memory task are presented, but the subject receives no instruction about the nature of the task. In the detection task, the subject must press a pushbutton as soon as a letter is presented. In the 1-back and 2-back tasks, the subject must decide whether the letter is identical to the one presented one trial back (1-back task) or two trials back (2-back task). Stimulus duration ¼ 500 ms, interstimulus interval (ISI) ¼ 5 s Four tasks were used. In the detection task, sequential letters (target) or background patches without letters (non-target) were presented. In the 1-back task, the target was any letter identical to the one immediately preceding it. In the 2-back task, the target was any letter that was identical to the one presented two trials back. Thus, working memory load increased from detection (memory-free) to 1-back (moderately demanding) and 2-back (highly demanding). In the ive fixation task, letter series identical to the 2-back task were presented, but subjects were unaware of the nature of the task and watched the series ively. Each task was tested in three stimulus sequences, composed of 30 images each (7 targets), adding up to 90 trials per task (21 targets). Before each sequence, subjects were informed about the nature of the forthcoming task. The protocol started with three sequences of the ive fixation task. A first detection task sequence was followed by a sequence of the 1-back task, three sequences of the 2-back task, two sequences of the 1-back task, and two sequences of the detection

EEG data were analyzed using NeuroScan software (NeuroScan Inc., Herndon, VA, USA). EEG signals were corrected for ocular artifacts using a threshold reduction algorithm (NeuroScan Inc.). The signal was digitally filtered in the 4–7.5 Hz theta frequency band, using narrow band filter (48dB=octave). The EEG data were automatically scanned for contamination by muscular or electrode artifacts, and visually inspected to control for other artifacts. EEG data were split into epochs of 4500 ms, starting 2500 ms before stimulus onset. The filtered epochs were then squared to obtain absolute spectral power magnitude, expressed in mV2. Theta power data were analyzed from trials with correct responses only. Due to the reduced number of response trials and to increase the signal-to-noise ratio, response and no-response trials were averaged, after controlling that there was no significant difference in theta power between them. Taking into its widespread scalp distribution, theta power was initially measured for each subject in the following regions: frontal (F4, Fz, F3), central (C4, Cz, C3), parietal (P4, Pz, P3), and occipital regions (O2, Oz, O1). Then, data from the three electrodes of each region were averaged. Within each task, the culmination peak of the theta event-related synchronization (ERS) was measured in the 0–500 ms time window after stimulus onset and the ERS magnitude was normalized to for inter-electrode and inter-subject variations in absolute values. Normalized ERS magnitude was computed as the percentage of the difference between the mean power during a chosen reference period (Baseline; 1100 ms to 100 ms relative to stimulus onset) and the ERS maximal value in the 0 to 500 ms post stimulus

1480

P. Missonnier et al.

interval (Maximum), according to the following formula: %ERS ¼ ððMaximum BaselineÞ=BaselineÞ 100

the 1-back and detection tasks (Fisher LSD: p<0.04 and p<0.01, respectively), while there was no significant difference between 1-back and detection tasks (Table 1).

Statistical analysis Differences in behavioral performance were analyzed with the Wilcoxon non-parametric test. Reaction time (RT) differences were assessed using a one-way repeated measures ANOVA, with task (ive, detection, 1-back, 2-back) as within-subject factor. A logarithmic (log) transformation was performed on theta %ERS in order to obtain a gaussian distribution of the values. Averaged culmination peak and log %ERS data over frontal, central, parietal and occipital electrode locations were used for statistics. Statistical analysis was performed on ERS culmination peak and log %ERS values separately using a two-way repeated measures ANOVA, with task (ive, detection, 1-back, 2-back) and electrode site (Frontal, Central, Parietal, Occipital) as within-subject factors. Post-hoc analysis was performed using Fisher’s Least Square Difference (LSD) test (Milliken and Johnson, 1984). Statistical threshold was set at p< 0.05.

Results Behavioural results Subjects performed very satisfactorily in n-back and detection tasks, with a minimum of 93% of correct responses (Table 1). Wilcoxon non-parametric test revealed that their performance was not significantly affected by the nature of the task. Repeated measures ANOVA showed highly significant task effects on reaction time (F(2, 22) ¼ 9.21, p<0.001). Post-hoc analysis revealed that reaction times in the 2-back task were significantly longer than in Table 1. Mean performance and reaction time standard error (n ¼ 12) Tasks

Performance

Reaction time

Detection 1-back 2-back

99.44 0.17 99.17 0.24 98.05 0.58

642.40 24.58 700.49 36.01 811.35 59.60;

Fisher LSD for reaction time; p < 0.05: 2-back vs. 1-back; p< 0.01: 2-back vs. detection

Electrophysiological results Theta power culmination peak and amplitude did not differ between response and noresponse trials, and these trials were thus pooled for further analysis. Within the reference period, the event-related absolute EEG theta power was low and constant in amplitude, and was not modulated by the type of task (Fig. 2). After stimulus presentation, a phasic increase of theta power (ERS) was observed during 0 to 500 ms for all tasks. Theta ERS culmination peak data are presented in Table 2. The 2-way repeated measures ANOVA revealed a main effect of electrode site on theta culmination peak (F(3, 33) ¼ 12.43, p<0.001), with theta ERS culminating earlier over occipital and parietal sites compared to frontal (Fisher LSD: p<0.001 and p<0.02, respectively) and central sites (Fisher LSD: p<0.0001 and p ¼ 0.05, respectively). No main task effect or interaction between task and electrode site was observed. While theta power amplitude preceding stimulus onset was equally low in all tasks, theta ERS immediately following stimulus onset was modulated in amplitude according to the task (F(3, 33) ¼ 5.74, p<0.003) but not to electrode site. However, there was a significant interaction between these two factors (F(9, 99) ¼ 2.99, p<0.003). Taskdependent modulations of the phasic theta %ERS were observed at frontal and central (p<0.001 and p<0.02 respectively) but not at parietal and occipital electrode sites. Figure 3 shows the mean theta %ERS values over the four electrode location sites for each task, and illustrates the statistical findings. Post-hoc analysis revealed that on frontal electrodes, theta %ERS was smaller in the ive fixation task as compared to each

Frontal theta event-related synchronization

1481

Fig. 2. Time course of mean event-related synchronisation (ERS) for the 4–7.5 Hz EEG frequency band as a function of task and electrode site (n ¼ 12). The y-axis depicts average theta power in mV2

1482

P. Missonnier et al. Table 2. Mean theta ERS culmination peak (ms) standard error (n ¼ 12)

Tasks

Frontal

Central

Parietal

Occipital

Detection 2-back 1-back ive fixation

276.57 17.97 268.27 11.83 261.69 17.92 239.05 20.95

253.43 17.91 258.94 13.17 264.19 22.27 234.57 22.19

225.72 24.24 224.05 24.18 222.73 18.08 189.40 18.37

203.33 23.63 185.00 24.76 197.32 20.33 192.93 16.38

Fig. 3. Mean theta %ERS value ( standard error) for each electrode site and task (n ¼ 12). Asterisks denote significant differences in post hoc tests (Fisher LSD,; p < 0.05; p < 0.01; p < 0.0001)

active task (Fisher LSD, detection vs. ive fixation, p<0.0001; 1-back vs. ive fixation, p<0.01; 2-back vs. ive fixation, p<0.05). This was also the case on central electrodes in comparison with detection task only (Fisher LSD, detection vs. ive fixation, p<0.006). Additionally, theta %ERS was of larger amplitude in detection as compared to 2-back task on frontal (Fisher: detection vs. 2-back, p<0.01) and central electrodes (Fisher LSD: detection vs. 2-back, p<0.004). No significant differences were obtained between de-

tection and 1-back task, or between 1-back and 2-back tasks. Discussion The role of theta oscillations in memory function has been largely debated since the discovery of hippocampal theta activity in rodents (Wetzel et al., 1977). In humans, the notion that task-related theta activity could be related to memory processing has been suggested after theta power increase was observed during memory tasks involving

Frontal theta event-related synchronization

storage and recall (Klimesch et al., 1994). However, working memory tasks also involve a major attentional component, and it remains unclear to which extent theta activity reflects an attentional framework for memory integration, rather than pure working memory processes. We attempted to resolve this issue by deg appropriate experimental conditions to dissociate the contribution of focused attention and working memory load to the genesis of theta activity. In addition, we introduced a fixation task involving minimal focused attention and cognitive demand, in order to evaluate the incidence of ive visual stimulation on theta activity. Performance was very close across tasks, and thus a similar number of correct responses was analyzed for each task. Furthermore, no significant difference in theta ERS was observed between response and noresponse trials, confirming the absence of theta reactivity to cerebral motor processes related to memory matching. Consequently, we opted to pool both types of trials for analysis in order to increase our signal to noise ratio. Theta power was of similar low level for all tasks in the reference period preceding stimulus presentation, suggesting that although tasks varied in attention and memory content, they involved a comparable amount of neuronal assemblies oscillating at theta frequency before visual stimulation. Phasic theta ERS started briskly with stimulus onset over all electrode sites, reaching a maximum at frontal and central electrodes within 500 ms after stimulus. This observation is concordant with previous observation reporting stimulus phase-locked theta activity turning on during cognitive task and off between trials, and culminating early after stimulus presentation (Caplan et al., 2001; Kahana et al., 2001; Raghavachari et al., 2001). The comparison of theta ERS between tasks led to three main findings. First, frontal phasic theta ERS was found of smaller amplitude in the ive fixation task as compared to the two memory tasks and the de-

1483

tection task. This observation s the hypothesis of a link between task demand and amplitude of frontal theta power. Phasic theta synchronisation emerging in relation to task demand s the role of theta activity in the integration of the incoming visual information for further cognitive process (Bastiaansen and Hagoort, 2003). Second, the amplitude of theta ERS was larger at frontal and central electrodes during the detection task as compared to the 2-back task, suggesting that attentional rather than memory processes contribute to enhance theta activity. Third, we did not observe any significant difference in theta power between 2-back and 1-back tasks, with theta ERS tending to be of smaller amplitude in the 2-back task. This result contrasts with the study of Krause and colleagues who reported a greater theta ERS at anterior electrode sites in a visual 2-back letter task compared to a 1-back task (Krause et al., 2000). An important methodological difference relative to the analysis frequency range might explain such discrepancy. While we chose to analyze the whole theta 4–7.5 Hz band, their results were obtained in the upper 6–8 Hz theta band, and observation of their figure reveals strong similarities between upper theta and lower alpha power time course. Lower alpha band ERD is generally itted to reflect general task demand, and showed a stronger reactivity to the 2-back task in Krause’s study. Our results suggesting that theta activity primarily reflects attentional processes, it is legitimate to wonder which neuroanatomical substrate could subtend this widespread theta oscillatory activity. Pioneer studies on cerebral attention networks in non-human primates may bring us some clues (Baleydier and Mauguiere, 1985; Mesulam, 1981). The parietal gyrus, prefrontal cortex, and cortical limbic structures (Mesulam, 1981) were demonstrated to participate to a distributed network of directed attention, together with the medial pulvinar nucleus (Baleydier and Mauguiere, 1985), which would play a key

1484

P. Missonnier et al.

role in interconnecting the cerebral structures (Baleydier and Mauguiere, 1985). Indeed, this nucleus projects to and receives inputs from the cortical structures of the network. Most importantly, it receives visual inputs from occipital cortex and superior colliculus, and connects to the parahippocampal structures (Baleydier and Mauguiere, 1985, 1987), forming an essential link with hippocampus for further memory processing (Van Hoesen and Pandya, 1975; Van Hoesen et al., 1979). The early rise and widespread distribution of theta activity following visual stimulation would be consistent with the activation of the directed attention network via extrastriate pathways, reaching the pulvinar through the superior culliculus and being subsequently distributed to all components of the network. While it cannot be ruled out that the network for directed attention might be activated after visual information has reached the striate cortex, the fast increase of theta power shortly after visual stimulus better fits with the former hypothesis. However, the high filters used for theta frequency band analysis may produce a shift of theta activity onset, and it is thus difficult to definitely conclude on the anatomical pathways involved. Our data the hypothesis that phasic theta activity may reflect the activation of the distributed network for directed attention within the extrapersonal space. Visual stimuli would activate the network as a function of task characteristics. Low level attention tasks, such as ive fixation task, would require fewer network resources and result in weaker theta response. Tasks demanding more attention would recruit enlarged oscillating neuronal populations, resulting in enhanced theta response. The finding of smaller theta ERS amplitude during the 2-back working memory task as compared to the detection task could be interpreted as a reallocation of resources for memory functioning, resulting in a net reduction of theta activity. Indeed, cognitive theories of working memory underline the interdependence of

attention and working memory processes (Baddeley, 1981, 1992), which have been shown to share in part common cerebral networks (LaBar et al., 1999). We employed the conventional eventrelated band power measure, which does not differentiate between evoked and induced components. The evoked component corresponds to the EEG power phase-locked to the stimulus, while the induced component refers to oscillations modulated by the stimulus but not responding in a phase-locked manner (Klimesch, 1999; Tallon-Baudry and Bertrand, 1999). The characteristics of theta ERS recorded in the present paradigm suggest the contribution of both evoked and induced components. Theta ERS recorded over the parieto-occipital regions culminated earlier than fronto-central theta ERS, and did not reveal any task-related effect. This posterior theta activity is likely to have close correspondence with evoked visual responses, as recently demonstrated by Klimesch and colleagues (Klimesch et al., 2004). In contrast, the task dependency of the fronto-central theta ERS argues for the different nature of this activity. This hypothesis is ed by previous EEG studies indicating an independence of frontal and parietal cortical brain regions to the genesis of the theta ERS response during cognitive tasks (Bastiaansen et al., 2002; Jensen and Tesche, 2002; Raghavachari et al., 2001; Sederberg et al., 2003). However further methodological investigation is necessary to determine whether frontal theta ERS is indeed dominated by induced oscillations. In conclusion, our results suggest that the early theta power increase during working memory tasks is not mainly related to longterm encoding and=or retrieval, but rather represents an EEG correlate of focused attention to target stimuli. Acknowledgements This project was funded by the Swiss National Foundation for Scientific Research, grant 3100-59110.99 (PG),

Frontal theta event-related synchronization the Research and Development HUG grant 03-I-15 (VI) and the Jer^ ome Tissieres Foundation, Stiftung f€ur Klinische Neuropsychiatrische Forschung and Ernst and Lucie Schmidheing Foundation (PG).

References Baddeley A (1981) The concept of working memory: a view of its current state and probable future development. Cognition 10: 17–23 Baddeley A (1992) Working memory. Science 255: 556–559 Baleydier C, Mauguiere F (1985) Anatomical evidence for medial pulvinar connections with the posterior cingulate cortex, the retrosplenial area, and the posterior parahippocampal gyrus in monkeys. J Comp Neurol 232: 219–228 Baleydier C, Mauguiere F (1987) Network organization of the connectivity between parietal area 7, posterior cingulate cortex and medial pulvinar nucleus: a double fluorescent tracer study in monkey. Exp Brain Res 66: 385–393 Basar-Eroglu C, Basar E, Demiralp T, Schurmann M (1992) P300-response: possible psychophysiological correlates in delta and theta frequency channels. A review. Int J Psychophysiol 13: 161–179 Bastiaansen M, Hagoort P (2003) Event-induced theta responses as a window on the dynamics of memory. Cortex 39: 967–992 Bastiaansen MC, van Berkum JJ, Hagoort P (2002) Event-related theta power increases in the human EEG during online sentence processing. Neurosci Lett 323: 13–16 Caplan JB, Madsen JR, Raghavachari S, Kahana MJ (2001) Distinct patterns of brain oscillations underlie two basic parameters of human maze learning. J Neurophysiol 86: 368–380 Demiralp T, Basar-Eroglu C, Rahn E, Basar E (1994) Event-related theta rhythms in cat hippocampus and prefrontal cortex during an omitted stimulus paradigm. Int J Psychophysiol 18: 35–48 Doppelmayr M, Klimesch W, Schwaiger J, Stadler W, Rohm D (2000) The time locked theta response reflects interindividual differences in human memory performance. Neurosci Lett 278: 141–144 Gevins A, Smith ME, McEvoy L, Yu D (1997) Highresolution EEG mapping of cortical activation related to working memory: effects of task difficulty, type of processing, and practice. Cereb Cortex 7: 374–385 Jensen O, Tesche CD (2002) Frontal theta activity in humans increases with memory load in a working memory task. Eur J Neurosci 15: 1395–1399 Kahana MJ, Seelig D, Madsen JR (2001) Theta returns. Curr Opin Neurobiol 11: 739–744

1485

Klimesch W (1999) EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brain Res Brain Res Rev 29: 169–195 Klimesch W, Doppelmayr M, Pachinger T, Ripper B (1997) Brain oscillations and human memory: EEG correlates in the upper alpha and theta band. Neurosci Lett 238: 9–12 Klimesch W, Schack B, Schabus M, Doppelmayr M, Gruber W, Sauseng P (2004) Phase-locked alpha and theta oscillations generate the P1-N1 complex and are related to memory performance. Brain Res Cogn Brain Res 19: 302–316 Klimesch W, Schimke H, Schwaiger J (1994) Episodic and semantic memory: an analysis in the EEG theta and alpha band. Electroencephalogr Clin Neurophysiol 91: 428–441 Krause CM, Sillanmaki L, Koivisto M, Saarela C, Haggqvist A, Laine M, Hamalainen H (2000) The effects of memory load on event-related EEG desynchronization and synchronization. Clin Neurophysiol 111: 2071–2078 LaBar KS, Gitelman DR, Parrish TB, Mesulam M (1999) Neuroanatomic overlap of working memory and spatial attention networks: a functional MRI comparison within subjects. Neuroimage 10: 695–704 McEvoy LK, Pellouchoud E, Smith ME, Gevins A (2001) Neurophysiological signals of working memory in normal aging. Brain Res Cogn Brain Res 11: 363–376 Mesulam MM (1981) A cortical network for directed attention and unilateral neglect. Ann Neurol 10: 309–325 Milliken GA, Johnson DE (1984) Designed experiments, vol 1. Van Nostrand Reinhold, New York O’Keefe J (1993) Hippocampus, theta, and spatial memory. Curr Opin Neurobiol 3: 917–924 Pfurtscheller G (1992) Event-related synchronization (ERS): an electrophysiological correlate of cortical areas at rest. Electroencephalogr Clin Neurophysiol 83: 62–69 Pfurtscheller G (2001) Functional brain imaging based on ERD=ERS. Vision Res 41: 1257–1260 Pfurtscheller G, Aranibar A (1977) Event-related cortical desynchronization detected by power measurements of scalp EEG. Electroencephalogr Clin Neurophysiol 42: 817–826 Raghavachari S, Kahana MJ, Rizzuto DS, Caplan JB, Kirschen MP, Bourgeois B, Madsen JR, Lisman JE (2001) Gating of human theta oscillations by a working memory task. J Neurosci 21: 3175–3183 Sederberg PB, Kahana MJ, Howard MW, Donner EJ, Madsen JR (2003) Theta and gamma oscillations during encoding predict subsequent recall. J Neurosci 23: 10809–10814

1486

P. Missonnier et al.: Frontal theta event-related synchronization

Tallon-Baudry C, Bertrand O (1999) Oscillatory gamma activity in humans and its role in object representation. Trends Cogn Sci 3: 151–162 Van Hoesen G, Pandya DN (1975) Some connections of the entorhinal (area 28) and perirhinal (area 35) cortices of the rhesus monkey. I. Temporal lobe afferents. Brain Res 95: 1–24 Van Hoesen GW, Rosene DL, Mesulam MM (1979) Subicular input from temporal cortex in the rhesus monkey. Science 205: 608–610 Wetzel W, Ott TA, Matthies H (1977) Post-training hippocampal rhythmic slow activity (‘‘theta’’) eli-

cited by septal stimulation improves memory consolidation in rats. Behav Biol 21: 32–40 Winson J (1978) Loss of hippocampal theta rhythm results in spatial memory deficit in the rat. Science 201: 160–163 Author’s address: P. Missonnier, Department of Psychiatry, Neuroimaging Unit, University of Geneva Hospitals, Ch. du Petit-Bel-Air 2, 1225 Geneva, Switzerland, e-mail: Pascal.Missonnier-Evrard@ hcuge.ch