P.H. Boeijinga

Interdependence of EEG signals: linear vs. nonlinear associations and the significance of time delays and phase shifts.

SummaryTo investigate the degree of interdependence of EEG signals, we have to use signal analysis methods. Three of these are described and their performance is compared: the cross-correlation (coherence and phase), the average amount of mutual information (AAMI) or the normalized AAMI, also called transmission coefficient T, and the correlation ratio h2 that is a general measure of nonlinear fit between any two signals. The three methods were applied to simulated and real signals in order to put in evidence how nonlinear relationships may affect differently these three measures of association. The nature of the interdependence between EEG signals is not characterized only by the degree of association, but also by the corresponding phase relationship. A basic question is whether such a phase shift can be interpreted as a transmission delay. However, a fundamental problem is that a phase shift may be difficult to interpret in terms of a biophysical model. A procedure is described in order to solve this problem. This involves computing the phase spectrum between the pair of signals, estimating the gain of the corresponding linear transfer function and the associated minimum phase. By subtracting the minimum phase from the phase spectrum, a corrected phase function can be obtained. From the slope of this phase function, a transmission delay can be estimated. This procedure is illustrated by applications to simulated and real EEG signals. It is demonstrated that from phase shifts we may estimate transmission delays between at least certain classes of EEG signals. In this way we can asses, unambiguously, how the transmission of information between different brain sites develops.

The spectral properties of hippocampal EEG related to behaviour in man.

In an epileptic patient with depth electrodes placed in the hippocampal formation we observed the following: (1) There existed a significant peak in the theta band of the hippocampal EEG of this subject under a number of behavioural circumstances. (2) Four different behaviours were performed at two degrees of intensity. Whenever the values of a certain spectral parameter associated with a low or a high intensity behaviour showed a significant difference, the higher value corresponded with the more intense version of the behaviour. (3) The frequency and rhythmicity of the hippocampal theta component during writing was consistently higher than during sitting or walking. The amplitude was smaller. (4) In a word association task the amplitude, frequency and rhythmicity showed a significant rise during the period of silence immediately following the question and preceding the answer.

Contribution of NMDA receptors to postsynaptic potentials and paired-pulse facilitation in identified neurons of the rat nucleus accumbens in vitro

SummaryThe principal aim of this study was to characterize the transmitter mechanisms mediating fast postsynaptic potentials in identified neurons of the rat nucleus accumbens. Using the biocytin-avidin labeling technique, impaled neurons were identified as medium spiny neurons. The basic membrane characteristics of these neurons were determined. Local electrical stimulation or stimulation of the corpus callosum elicited a depolarizing postsynaptic potential consisting of an EPSP often followed by an IPSP. The quisqualate/kainate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (4 μM) abolished most of the depolarizing postsynaptic potential. The N-methyl-D-aspartate receptor antagonist D(-)-2-amino-5-phosphonopentanoic acid depressed a small part of the decay phase of the depolarizing postsynaptic potential. Paired-pulse facilitation of postsynaptic potentials was found using interstimulus-intervals between 10 and 150 ms. N-methyl-D-aspartate receptors were found to contribute only slightly to the facilitation of the decay phase of the depolarizing postsynaptic potential, but not to its rising phase. This contribution was particularly clear under conditions of reduced GABAA receptor mediated inhibition. The present study indicates that postsynaptic responses of medium spiny neurons in the nucleus accumbens to local stimulation or stimulation of neocortical afferents are primarily mediated by quisqualate/kainate receptors. The contribution of NMDA receptors is normally limited to a portion of the decay phase of these responses, but is enlarged in the absence of GABAergic inhibition and following paired-pulse stimulation.

Locally evoked potentials in slices of the rat nucleus accumbens: NMDA and non-NMDA receptor mediated components and modulation by GABA

In a slice preparation of the rat nucleus accumbens (Acb), local electrical stimulation elicited a field potential composed of two negative peaks, followed by a positive wave. The early negative peak was identified as a non-synaptic compound action potential, the late negative peak as a monosynaptic population spike (PS) and the positive wave as a mixture of an excitatory and an inhibitory postsynaptic potential (PSP). Both the PS and the PSP exhibited a marked degree of paired-pulse facilitation. The quisqualate/kainate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 2 microM) and the broadly acting glutamate receptor antagonist kynurenic acid (300 microM) reversibly abolished or reduced both the PS and PSP. In contrast, nicotinic, muscarinic and N-methyl-D-aspartate (NMDA) receptor antagonists had no suppressive action. Washout of Mg2+ from the superfusion medium reversibly enhanced and prolonged the PSP and this effect was blocked by the NMDA receptor antagonist D(-)-2-amino-5-phosphonopentanoic acid (D-AP-5). The gamma-aminobutyric acid antagonist picrotoxin (60 microM) enhanced the PS and induced secondary spikes which were superimposed on a prolonged PSP. Most of this prolongation was abolished by D-AP-5. It is concluded that locally evoked synaptic responses in the Acb are mediated by glutamate or aspartate, and that NMDA receptor mediated activity evoked by low frequency stimulation is substantial in Mg2(+)-free medium or during reduced GABAA receptor activity, but not under normal conditions.

Responses of the nucleus accumbens following fornix/fimbria stimulation in the rat: identification and long-term potentiation of mono- and polysynaptic pathways

The nucleus accumbens occupies a strategic position as an interface between limbic cortex and midbrain structures involved in motor performance. The fornix-fimbria carries limbic inputs to the ventral striatum, namely by way of fibers originating in the CA1/subiculum and projecting to the nucleus accumbens. It also carries fibers arising in the septal area that project to the hippocampal formation, and projection fibers to other areas of the rostral forebrain from Ammons horn. Electrical stimulation of this bundle causes characteristic field potentials both in the nucleus accumbens and in the subiculum. In rats, under halothane anesthesia, the responses evoked by fornix/fimbria stimulation in the nucleus accumbens consist of two main positive peaks (at 10 and 25 ms, referred to as P10 and P25, respectively). P10 represents monosynaptic activation. We hypothesized that P25 reflects the activation of a polysynaptic loop, i.e. a fornix-fimbria hippocampal loop in series with the fibers that arise in the subiculum and project to the nucleus accumbens. To test this hypothesis, we reversibly blocked the fibers projecting caudally to the hippocampus by a local anesthetic (lidocaine) and the glutamatergic transmission through the CA1/subiculum by a local injection of kynurenic acid. Both manipulations yielded a reversible depression of about 90% of the P25 component while P10 remained unaffected as expected. In concert a strong reduction (to 24-31%) of control values of the responses evoked in the subiculum was seen. The dynamics of the mono- and polysynaptic pathways differ markedly. The synaptic responses through both pathways are enhanced by paired-pulse stimulation, but the polysynaptic pathway is facilitated in a much stronger way. Following a tetanus (50 Hz, 2 s duration) applied to the fornix/fimbria, the P10 component of the nucleus accumbens responses showed an immediate increase by a factor of about 2 followed by a phase of gradual decrement with half-decay time of about 10 min, after which a persistent long-term potentiation of about 25% above control level was maintained for the rest of the experiment (max 90 min). The P25 component showed a transient 10-fold potentiation with return to control values after about 10 min. In contrast to the P25 elicited by a conditioning stimulus, the P25 component elicited by a second stimulus delivered at an interval of 100 ms (test stimulus) showed a persistent long-term potentiation.(ABSTRACT TRUNCATED AT 400 WORDS)

Modulations of EEG activity in the entorhinal cortex and forebrain olfactory areas during odour sampling

With the aim of determining a possible role of the entorhinal cortex (EC) in the processing of olfactory information, its electrical activity was recorded during different types of odour sampling behaviour. The EC spontaneous field potentials (EEGs) were related to those recorded simultaneously from the more rostrally lying olfactory bulb (OB) and prepiriform cortex (PPC), using coherence functions. The coherence values are measures of coupling between two neuronal populations. Auto- and coherence spectra were characterized by a peak in the beta-band (35-36 Hz) and another in the 16-20 Hz band. For the beta-peak of the PPC-EC coherence it was shown that both the maximal values and the median frequency decrease during the first 0.5-1.0 s of a trial in a two-choice odour discrimination task. These transients differed significantly for CS+ and CS- trials. However, no such difference was observed during exploratory sniffing at the same pair of odorants. It was concluded that during sniffing the degree of coupling of the EC with the PPC depends on the behavioural context but not on the quality of the odorants. As regards the 16-20 Hz components it was shown, using bispectral analysis, that these components represent the 1/2 subharmonic of the beta-components. The subharmonic components were enhanced during the initiation of sniffing. This means that the neural networks of OB, PPC and EC have non-linear dynamic properties. These networks show different modes of oscillatory behaviour, characteristic of the restful state and of the active sniffing state. Theoretical implications of these experimental results are discussed. In conclusion, it was demonstrated that modulations with sniffing behaviour are present in the beta-EEG components and their subharmonics in the olfactory brain areas including the EC. This indicates that the latter is involved in processes of the evaluation of olfactory cues in relation to the animals behaviour.

Differential distribution of β and θ EEG activity in the entorhinal cortex of the cat

The basic characteristics of the electroencephalograms (EEGs) of the entorhinal cortex (EC) were compared to those of the olfactory bulb (OB) and prepiriform cortex (PPC) in the awake cat, using methods that take into account linear (coherence functions) and non-linear relationships (mutual information). The frequency spectra of EC signals differ from those of EEG signals recorded from OB or PPC, but in the beta frequency band (35-40 Hz) coherent activity between these latter two areas and the EC could be put in evidence. Taking the values of coherence as a measure of linear coupling between populations of neurons, this coupling is strong between PPC and the rostral part of the EC, and becomes weaker towards more lateral, caudal and medial areas. This decrease in coupling is confirmed by non-linear analysis methods. On the basis of intracortical depth profiles it was shown that the beta activity is locally generated in layer II of the EC. Another rhythmic activity was also recorded in EC in the theta frequency range. Theta rhythm is generated mainly in the caudal part of EC. This area has the strongest coupling with the hippocampal formation. These analyses lead to the conclusion that in the EC olfactory and limbic domains have different topographical distributions but with a certain degree of overlap.

Paired-pulse facilitation in the nucleus accumbens following stimulation of subicular inputs in the rat

Anatomical tracing studies indicate that the nucleus accumbens receives inputs from limbic structures, and projects to the ventral pallidum. In order to get more fundamental insight into how information from the limbic areas is relayed via the nucleus accumbens, electrophysiological experiments were carried out in rats under halothane anaesthesia. Inputs originating in the subiculum were activated by electrical stimulation of the fornix fibres, and both field potentials and extracellular unit activity were recorded from the medial and lateral aspects of the nucleus accumbens. Evoked potentials consisted of two positive peaks (P1 at 10 ms and P2 at 25-30 ms). In between a negative-going wave (N1) was present. These initial components were followed by a complex negative wave (N2) with variable duration of 30-100 ms. The P2 and N2 components showed a conspicuous paired-pulse facilitation at stimulus intervals between 80 and at least 200 ms. When responses were recorded at increasing stimulus intensity, the second response emerged at lower threshold than the first response. The mechanisms underlying these phenomena were investigated by analysing the extracellularly recorded unit activity. Primarily, excitatory responses were found. Onset-latencies could be divided roughly into two clusters, one around 10 ms, representing monosynaptic inputs, and a second around 24-26 ms. Inhibitory responses were also found. Stimulation of the ventral pallidum was carried out in order to test whether the cells that could be driven by stimulation of the subicular inputs were projection cells. Latencies of antidromic action potentials ranged from 9 to 13 ms. A minority of the identified projection cells were activated by limbic inputs. The projection cells were found in the core region of the nucleus accumbens. Units that were inhibited by stimulation of the limbic inputs were found in the shell only, whereas excitatory responses were measured in both subdivisions of the nucleus accumbens. For the latter responses a significant enhancement, by a factor of four, was found using double pulse stimulation of the fornix at intervals of 100 ms. The basic electrophysiological properties are compared with those described in the literature, and speculations about the possible mechanisms responsible for the paired-pulse facilitation phenomena are put forward.

Hippocampal EEG and motor activity in the cat: The role of eye movements and body acceleration

In cat the relation between various behaviours and the spectral properties of the hippocampal EEG was investigated. Both EEG and behaviour were quantified and results were evaluated statistically. Significant relationships were found between the properties of the hippocampal EEG and motor acts (walking, sitting, eating, stepping and eye movements). These results were compared with those obtained in dog under similar experimental circumstances. Species differences were found particularly regarding the fact that in the cat a dissociation between frequency and amplitude parameters was obtained for some behaviours; this may explain why appreciable differences in the visual interpretation of EEG records of different species are often reported. A simple model of the modulation of hippocampal EEG by brainstem inputs is presented. Particular attention is paid to species differences regarding the anatomy and physiology of the pathways involved in this modulation. It is concluded that in cat a strong relation exists between the modulation of spectral properties of hippocampal EEG and vestibular inputs and/or eye movements. The effects of body acceleration on hippocampal EEG are put in evidence and related to theories of hippocampal function.

A new method to estimate time delays between EEG signals applied to beta activity of the olfactory cortical areas.

A method based on coherence and phase spectra was developed to characterize the transmission of signals through neural networks that were assumed to have linear filter properties. The system is thought to consist of a frequency independent delay in series with a network that gives rise to a frequency dependent lag, the so-called minimum phase shift. The latter was estimated from the gain of the transfer function by means of Hilbert transform pairs. By subtracting the minimum phase shift from the phase differences between input and output, a corrected phase spectrum was obtained that represents the frequency independent delay. The slope of this spectrum was calculated and converted into the time delay. This method was applied to EEG signals recorded from different parts of the olfactory basal forebrain structures of the cat, in order to determine time delays for the beta activity. The results showed that the beta activity was propagated in a rostro-caudal direction. Along this rostro-caudal axis the conduction velocity slowed down from approximately 3 m/sec for the pathways between olfactory bulb and prepyriform cortex to 0.5 m/sec within the EC. These findings were compared to estimates of the conduction velocities in the lateral olfactory tract fibres and collaterals using transient responses. It is concluded that, to a first approximation, the linear approach gives physiologically meaningful results.