A transient dominance of theta event-related brain potential component characterizes stimulus processing in an auditory oddball task
Introduction
Cellular, local field, and electroencephalographic (EEG) oscillations provide a link between brain activity and basic mechanisms of information processing (e.g. Singer, 1998, Basar, 1998). Following external stimulation, functionally meaningful EEG responses from multiple frequency ranges known as event-related oscillations (gamma, 30–70 Hz; alpha, 7–14 Hz; theta, 4–7 Hz; delta, 0.1–4 Hz) can be revealed by decomposing the compound EEG signal (Basar, 1998, Basar, 1999). In the framework of event-related brain dynamics, the functional reactivity of oscillations from a given frequency range may depend on and/or covary with activations from other frequency channels. Thus, both global and local interactions of multiple frequencies can further disclose the neuroelectric functional involvement of oscillations (Friston, 1997, Nunez, 2000, Bressler and Kelso, 2001).
Although in the majority of studies in humans, processes at the same frequency are dealt with, an interdependence between multiple oscillations has also been demonstrated. For example, concerning temporal relationships, the timing of evoked (stimulus-locked) theta oscillations elicited by rare tones predicts the latency of P300-delta component of event-related potentials (ERPs) (Yordanova and Kolev, 1997, Yordanova and Kolev, 1998a, Yordanova and Kolev, 1998b), and this delta component may trigger subsequent event-related desynchronization of alpha frequency (7–14 Hz) components (Yordanova and Kolev, 1998c, Yordanova et al., 2001). Likewise, the onset of fast alpha desynchronization correlates with the timing of theta activity distribution during memory retrieval (Sauseng et al., 2002). Further, evoked gamma oscillations precede beta oscillations in response to auditory stimulus suggesting a gamma-to-beta EEG transition (Haenschel et al., 2000). Spatial relationships between multiple frequencies have also been established and functionally linked with neural coding (Friston, 1997, Basar, 1998). Accordingly, neuromagnetic gamma oscillations in the prefrontal cortex have been associated with beta oscillations in the parietal cortex in humans (Friston, 1997).
Interactions among temporally and spatially overlapped multiple frequency components are of particular interest, as exemplified by oscillatory models according to which dual (theta and gamma) oscillations can underlie short-term memory (Lisman and Idiart, 1995, Schack et al., 2002), or a fast information acquisition in neocortical–hippocampal circuits (Buzsaki, 1996). Moreover, functional ERP components from different frequency ranges may occur simultaneously (Basar, 1980, Basar, 1998, Kolev et al., 1997, Demiralp et al., 1999, Yordanova et al., 2000), but they are typically analyzed independent of each other and are correlated with sensory and cognitive processes in specific ways. It is least well known how temporally superimposed frequency ERP responses recorded from a given scalp location interact with each other under specific processing conditions. Therefore, the aim of the present study was to explore dynamic relationships of locally co-existent event-related oscillations during task processing.
One possible approach to address this issue is to analyze how the complex behavior of the EEG response is modified by multi-frequency relationships. An advantageous way to study complex EEG dynamics is to implement the entropy concept (Wright, 1999, Reinagel, 2000). According to information theory, entropy reflects states of order/disorder of a system (Shannon, 1948). When applied to Fourier transformed EEG, entropy can measure how concentrated or widespread the power spectrum of the EEG signal is (Inouye et al., 1991, Inouye et al., 1993). Low spectral entropy values correspond to a narrow-band (mono-frequency) activity characterizing highly ordered (regularized) bioelectric states, and high entropy values reflect a wide-band (multi-frequency) activity. The spectral entropy has certain limitations to detect ‘regularized’ microstates in short-duration EEG signals such as ERPs because of the low time resolution of the Fourier transform. To overcome these limitations, entropy based on time–frequency EEG decomposition by means of the wavelet transform (WT) has been proposed, called wavelet entropy (WE) (Blanco et al., 1998, Quian Quiroga et al., 2001, Rosso et al., 2001, Rosso et al., 2002). Since the WT represents each EEG frequency with near-optimal time resolution (e.g. Schiff et al., 1994, Ademoglu et al., 1997, Demiralp et al., 1999, Samar et al., 1999), the WE can quantify more precisely the time dynamics of EEG order/disorder. It is, therefore, proposed in the present study that a time-localized entropy decrease can reflect a transient dominance of one particular frequency ERP component over other frequency ERP components. Conversely, high WE values would correspond to a simultaneous change of multiple frequency ERP components in the same direction (i.e. a simultaneous increase or decrease). Thus, specific types of interrelations among co-existent oscillations from different frequency ranges can be analyzed as a function of time.
Recent applications of WE to ERPs have revealed a new bioelectric event that has not been described previously. This was a short-lasting dominance of synchronized theta (4–8 Hz) ERP responses over ERP responses from other frequency bands, which produced a time-localized and transient decrement of ERP entropy in the post-stimulus epoch (Rosso et al., 2001, Kolev et al., 2001, Yordanova et al., 2002). Specifically, a developmental model has shown that such a transient WE decrement (a highly ordered ERP microstate1) characterizes passive auditory ERPs of both children and adults (Kolev et al., 2001). Despite age-related variations in the frequency content of both the background and post-stimulus EEG, the WE decrement in all age groups resulted from a strong transitory dominance of synchronized theta activity. Further, in auditory, visual, and bisensory (audio–visual) processing conditions, a similar short-lasting WE decrease has been detected after stimulus from each modality such that this decrease was again determined by a dominant theta ERP component (Yordanova et al., 2002). Thus, a transient theta dominance consistently occurred after external stimulation in a manner that did not depend on developmental or modality-related EEG frequency characteristics.
Importantly, in contradistinction to power and/or synchronization changes of theta ERP components during task-stimulus processing described in other reports (e.g. Basar-Eroglu et al., 1992, Klimesch, 1999, Yordanova and Kolev, 1997, Yordanova and Kolev, 1998a, Yordanova and Kolev, 1998b, Basar, 1998, Caplan et al., 2001, Sarnthein et al., 1998, Gevins et al., 1997), theta dominance reflects the ratio or interdependence among locally co-existent multiple frequency ERP components rather than solely theta frequency dynamics. In this context, event-related theta dominance may represent a very different and integrative neurobioelectric phenomenon.
As mentioned, WE and theta dominance have been quantified in passive (no-task) conditions which do not elicit substantial endogenous ERP components (Rosso et al., 2001, Kolev et al., 2001, Yordanova et al., 2002). Passive stimulus evaluation is typically accompanied by theta, alpha and faster frequency responses (Basar, 1998, Yordanova and Kolev, 1997, Yordanova and Kolev, 1998a, Yordanova and Kolev, 1998b, Yordanova and Kolev, 1998c), whereas responses from the slow (delta) frequency range are enhanced by task demands and underlie the expression of late endogenous waves such as P300 (Basar-Eroglu et al., 1992, Kolev et al., 1997, Demiralp et al., 1999, Spencer and Polich, 1999, Yordanova et al., 2000, Schürmann et al., 2001). In such a case, due to the fact that lower frequencies of the EEG have larger amplitudes than higher frequencies, the so identified theta-dominated miscrostate in no-task ERPs may have been biased by the less pronounced slow endogenous components from the delta range. Accordingly, a recent study using a visual oddball task has suggested that the WE decrement observed in the post-stimulus period may be caused by the large P300-delta components generated in this condition (Quian Quiroga et al., 2001). Hence, the question arises whether the theta dominance detected in no-task experiments is restricted to passive stimulus processing or whether it would be also present in task-related ERPs comprising enhanced delta activity.
In the present study, this question was approached by analyzing WE minimum in an active (mental count) auditory oddball condition with elevated discrimination difficulty: (1) This condition is known to induce pronounced differences between target and non-target processing, most reliably reflected by P300-delta components (Polich, 1998). (2) Earlier analyses in the time–frequency domain have shown that the delta ERP components elicited in this task are substantially larger than those elicited in other types of oddball tasks (e.g. requiring motor response) (Kolev et al., 1997, Kolev et al., 2001, Yordanova and Kolev, 1998a). (3) Not to further favor the relative expression of theta ERP components and verify more reliably a presumable theta dominance, recordings were made with eyes closed in order to enhance the contribution of alpha activity in the ERPs.
Task effects were analyzed for the timing, amount, and frequency determinants of ERP WE minimum (Rosso et al., 2001, Rosso et al., 2002). ERPs were also measured in the time domain to demonstrate differences between task-relevant and task-irrelevant stimulus processing as reflected by ERP components, and P300 in particular. ERPs were further analyzed in the frequency domain to compare the spectral power of ERP components from delta, theta, alpha, and faster frequency ranges and explore whether WE parameters may depend on the frequency content of target and non-target EEG responses.
Section snippets
Subjects and recording conditions
EEG and ERPs were recorded from 10 right-handed healthy subjects (6 females, mean age=26.15 years, SD=3.28 years). In an oddball condition, a total of 100 auditory stimuli were presented with an intensity of 60 dB sound pressure level, duration of 1000 ms (radio frequency 10 ms), and inter-stimulus intervals between 3.5 and 6.5 s. Two stimulus types, targets (1950 Hz, P=0.20) and non-targets (2000 Hz, P=0.80), were randomly delivered. Targets had to be counted silently. During recordings subjects
Results
Eight out of 10 subjects performed the task accurately and reported the correct number of counted targets. Two subjects have counted one and two targets less, respectively.
Discussion
To assess whether and how locally co-existent frequency ERP components may depend on each other during task processing, the present study analyzed the behavior of the complex EEG response at specific electrodes in an auditory oddball task. Complex EEG dynamics were quantified by means of ERP entropy calculated as a function of time (Rosso et al., 2001, Rosso et al., 2002). Major results revealed that a highly ordered EEG microstate emerged in response to both target and non-target stimuli, as
Acknowledgements
This work was supported by grants to J.Y. and V.K. from the James S. McDonnell Foundation, USA (98-66 EE-GLO-04), the Deutsche Forschungsgemeinschaft, Germany (436-BUL-113/105), and the National Research Council at the Ministry of Science and Education, Bulgaria (B-812/98), as well as by grants to O.A.R. from the International Office of BMBF, Germany (ARG-4-G0A-6A), Consejo Nacional de Investigaciones Cientı́ficas y Técnicas (CONICET), Argentina (PIP 0029/98), and Fundación Alberto J. Roemmers,
References (56)
- et al.
The selectively distributed theta system: functions
Int J Psychophysiol
(2001) - et al.
Gamma, alpha, delta, and theta oscillations govern cognitive processes
Int J Psychophysiol
(2001) - et al.
P300-response: possible psychophysiological correlates in delta and theta frequency channels. A review
Int J Psychophysiol
(1992) - et al.
Cortical coordination dynamics and cognition
Trends Cogn Sci
(2001) - et al.
Time–frequency analysis of single-sweep event-related potentials by means of fast wavelet transform
Brain Lang
(1999) Another neural code?
Neuroimage
(1997)- et al.
Transformations towards the normal distribution of broad band spectral parameters of the EEG
Electroenceph clin Neurophysiol
(1982) - et al.
Quantification of EEG irregularity by use of the entropy of the power spectrum
Electroenceph clin Neurophysiol
(1991) - et al.
Localization of activated areas and directional EEG patterns during mental arithmetic
Electroenceph clin Neurophysiol
(1993) EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis
Brain Res Rev
(1999)
EEG alpha map series: brain micro-states by space-oriented adaptive segmentation
Electroenceph clin Neurophysiol
The dimensional complexity of the P3 event-related potential: area-specific and task-dependent features
Clin Neurophysiol
Wavelet entropy: a new tool for analysis of short time brain electrical signals
J Neurosci Methods
Time–frequency analysis of sensorial brain activity
Suppl Clin Neurophysiol
Wavelet analysis of neuroelectric waveforms: a conceptual tutorial
Brain Lang
The interplay between theta and alpha oscillations in the human electroencephalogram reflects the transfer of information between memory systems
Neurosci Lett
Phase-coupling of theta–gamma EEG rhythms during short-term memory processing
Int J Psychophysiol
Fast wavelet transformation of EEG
Electroenceph clin Neurophysiol
Delta responses and cognitive processing: single-trial evaluations of human visual P300
Int J Psychophysiol
Developmental changes in the relationship between EEG theta response and P300
Electroenceph clin Neurophysiol
Wavelet entropy analysis of event-related potentials indicates modality-independent theta dominance
J Neurosci Methods
Analysis of pattern reversal visual evoked potentials (PRVEP's) by spline wavelets
IEEE Trans Biomed Eng
EEG brain dynamics. Relation between EEG and brain evoked potentials
Brain function and oscillations
Brain function and oscillations
Time–frequency analysis of electroencephalogram series III. Wavelet packets and information cost function
Phys Rev E
Short duration power changes in the EEG during recognition memory for words and faces
Psychophysiology
The hippocampo-neocortical dialogue
Cereb Cortex
Cited by (39)
Protective effect of Devosia sp. ANSB714 on growth performance, serum chemistry, immunity function and residues in kidneys of mice exposed to deoxynivalenol
2016, Food and Chemical ToxicologyCitation Excerpt :Derivatives of DON include its deep oxidized form due to microbial metabolism, which is called de-epoxy-DON (Yoshizawa et al., 1983) or simply de-DON, and conjugated forms as a result of phase II metabolism, which happens mainly in intestinal mucosa, liver, and kidney (Dänicke and Brezina, 2013). Yordanova et al. (2003) used ELISA method to determine DON and found that after mice orally administered with 25 mg DON/kg b.w., DON concentrations were significantly elevated in kidney > heart > plasma > liver > thymus > spleen > brain from 30 min to 8 h, with a rapid clearance occurring over 24 h, but DON was undetectable except in kidney at 24 h. So, the kidney may be a major target tissue for residues of DON. Toxin diets supplemented with Devosia sp.
Gender modulates the development of theta event related oscillations in adolescents and young adults
2015, Behavioural Brain ResearchCitation Excerpt :Studies of visual and auditory target detection tasks using functional magnetic resonance imaging (fMRI) suggest that common, supramodal functional systems are involved as well as modality-specific systems [6,7]. Frequency domain analysis suggests that the theta band event related oscillation (ERO) is a major constituent of the P3 response [8–13]. Theta EROs are important for processes underlying frontal inhibitory control, conscious awareness, recognition memory and episodic retrieval, as shown in a number of experimental contexts [14–19].
The effect of aging on EEG brain oscillations related to sensory and sensorimotor functions
2014, Advances in Medical SciencesCitation Excerpt :The delay of theta amplitude reversal between the tones tasks, performed by OG, could be accounted for by the delay in P3 peak appearance after HT in OG. Previous findings have demonstrated that transient theta dominance characterized a stimulus modality, and constituted a particular microstate when stimulus information is transferred, integrated or maintained in working memory [42–44]. Widespread stronger theta activation was seen for YG after HT during the late cognitive processing.
Wavelet spectra of visual evoked potentials: Time course of delta, theta, alpha and beta bands
2013, NeurocomputingCitation Excerpt :It is important to note that the alpha band ERD should be considered not just a marker for some ERP activity, but plays an active role in cortical information processing during perception and memory [20,21]. An ERS in the theta band was reported during target detection in few studies [22,23]. Less prominent ERD and ERS processes may occur in the other EEG bands as well [5].