Oscillatory Brain Theory

Oscillatory Brain Theory: A New Trend in Neuroscience The Role of Oscillutory Processes in Sensory and Cognitive Functions ccording to Freeman [ 191 neuroscience is ripe for change. During the “Decade of the Brain,” brain science is coming to terms with its ultimate problem: understanding the mechanisms by which the immense number of neurons in the human brain interact to produce the higher cognitive functions. The analysis of the brain’s natural frequencies opens a new window toward a combined analysis of sensory and cognitive functions at the level of single neurons and the field potentials (EEG or MEG) [3,4]. In the last decade, our research group has been strongly involved in the development of nonlinear brain dynamics and with the oscillatory processes of neural assemblies. During this period, several conferences were organized (New York 1992, Liibeck 1994, Travemiinde 1996) and several books related to this trend were edited [6,7 1. Justrecently, three new volumes extensively described this new evolution in neuroscience by concluding that a new integrative neurophysiology and a new “brain theory” is needed in order to confront the problems recognized in this decade of the brain. This article provides a brief outline of how the findings in the last 20 years of research have led to such a development.

A

Erol Basar’J, Canan Basar-Eroglu3,Sirel Karakas2r4and Martin Schurmann’

’

Institute of Physiology, Medical University liibeck, Germony ’TUBITAK Broin Dynamics Research Unit, Turkey ’Institute of Psychology and Cognition Research, Germony 41nstituteof Experimental Psychology, Hacettepe University, Turkey

Developmentsin Oscillatory Brain Theory According to Freeman [18] the last revolution of ideas about the brain took place in the middle of the century now ending, when the field took a sharp turn into a novel direction. During the preceding five decades the prevailing view, carried forward from the 19th century, was that neurons are the carriers of nerve energy, either in chemical or electrical forms [ 181. Neuron populations interact with each other across extended regions of the brain by large bundles and tracts of axons. IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Each part of cortex and basal ganglia maintains its own “soap bubble” dynamics, with specializations based in its history and input, and it is pushed by these interactions into creating new patterns within itself that reflect and contiibute to an ever-shifting global pattern involving the entire forebrain. These patterns are not re-entrant “mappings” that correspond to transfers of information in computational neural networks. Rather, they are jynamical flows with continuous distributions and trajectories, comparable to humcanes and tornadoes. The mathematics needed to describe these flows has undergone striking developments in recent years with the aid of computer graphics and digital computers, particularly as ad.ipted by Abraham and Shaw [ 11 for nonspccialists. According to Freeman [ 181. this is the level of predominant concern a:, also in the long-standing studies of our gi’oup,extensively described by Basar [3, 41. According to our new scope [3,4] an “integrative neurophysiology” should describe activity of several brain areas and their multifold functions from 1 global viewpoint. Further, it must rely cn extensive experimental work: the EEC oscillations permit the analysis of sensory and cognitive functions together---both in freely behaving animals and in the thinking and feeling human brain. Prxessing of sensation and cognitit,e func, ions are interwoven; the type of approack utilized in this article makes the analysi: of integrative brain functions feasible. An integrative neurophysiology must embrace both physiologically vital and cognitive functions. The bridging of both types of functions is possible only by a ,ommon approach. Brain oscillations seem to be basic events for establishing such an integrative discipline. Since the nuturnl frequencies of the central nervous system (CNS) are manifested in EEG oscillations 0739-5175/99/$lO.OOC319991EEE

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

Ma//June 1999

(delta: approx. 0.5-3.5Hz; theta: approx. 4-7 Hz, alpha: approx. 8-13 Hz, and beta: approx. 15-30 Hz), such an approach is probably one of the most fruitful or promising approaches to creating integrative neuroscience for the future.

potentials (ERPs) are recorded in the following way: a defined event occurs several times (e.g., 100 sensory stimuli are presented; however, a wide variety of other events were used as well). With every event, a segment of the EEG preceding the event and a segment of the EEG following the event (the latter referred to as a single-trial, event-related potential) are digitized and stored. Averaging of the single-trial ERPs yields the (averaged) ERP. As to further evaluation, the experiments share a common approach, the “combined analysis procedure” [3]:

Event-Related Oscillations in Brain Function Most functions require the integrated action of neurons located in many regions. Localization of function means that certain areas of the brain are more concerned with one kind of function than with others. The existence of a significant difference in the major operating rhythms in occipital or Frequency-Domain Analysis of frontal areas gives strong support to the Averaged ERPs possibility that theta (spontaneous, Fourier transformation of the ERP time evoked, induced) and alpha rhythms series yields the amplitude frequency (spontaneous, evoked, induced) have fun- characteristics (AFCs). Additionally, the damentally different functional operations ERP time series are digitally filtered in [3,4,8]. But during some functional states, several frequency ranges defined accordmajor operating rhythms can change their ing to the AFC. functional roles; the nature of the experiment (i.e., tasks) can influence the weight Single Sweep Analysis of these functional components on brain In order to understand the physiological rhythms (see chapter 27 in [4]). or cognitive contents of ERPs, single triAt the turn of the 20th century, Karl als were analyzed using (a) pass-bandfilLashley believed that various parts of the tering, (b) wavelet analysis, and (c) brain were equipotential and that, for single-sweep wave identification for the many functions, virtually any part of the assessment of phase-locking single brain could substitute for any other. Con- sweeps independent of amplitude (see trary to this hypothesis, subsequent exper- chapters 4, 5 and 6 in [3]). By means of the application of comiments showed that even highly complex brain functions can be attributed to spe- bined analysis procedure of EEG and EPs, cific brain areas. Localization does not we recently emphasized the functional imply, however, that any specific function importance of oscillatory responses (in is exclusively mediated by only one re- the framework of brain dynamics) related gion of the brain. Following this line of to association and (“long distance”) comthought, several event-related oscillations munication in the brain. We assumed that in various frequency bands are, in the fol- alpha networks, theta networks, and lowing synopsis, assigned to multifold gamma networks (or systems) are selecbrain functions. Experiments hinting at tively distributed in the brain (for the functional correlates of event-related os- delta, theta, and alpha ranges see chapters cillations will be summarized below. The 24,25,26 in [4]). We also have tentatively results will be embedded into a concept of assigned functional properties, namely “oscillatory networks” or “oscillatory sensory-cognitive functions, to alpha, theta, delta, and gamma resonant resystems” in the brain. sponses. For example, a sensory stimulation evokes 10 Hz enhancements in Selectively Distributed Oscillatory several structures of the brain, both cortiSystems in the Brain Oscillatory responses (i.e., changes of cal (primary auditory cortex, primary viongoing EEG activity temporally related sual cortex) and subcortical to a defined event; e.g., a sensory stimu- (hippocampus) lus) have been observed in several strucThe synchronous occurrence of such tures of the brain, both cortical and responses in multiple brain areas hints at subcortical. The properties of such oscil- the existence of distributed oscillatory lations will be summarized below. Only a systems and parallel processing in the rough outline can be presented here (for brain. Such diffuse networks would facilidetails of the experiments, see the respec- tate the information transfer in the brain, tive references). Typically, event-related according to the general theory of resoMoyjJune 1999

nance phenomena. Although alpha responses are observable in multiple brain areas, they are markedly dependent on the site of recording. The dependence of the alpha response on whether or not the stimulus is adequate for the brain area under study thus hints at a special functional role of alpha responses in primary sensory processing. It is not yet not possible to define connections between the elements of these systems or to define the directions of signal flow and exact boundaries of neuronal populations involved. However, this description is necessary to emphasize that rhythmic phenomena in these frequency ranges are not unique features of the observed single subsystem of the brain, and that their simultaneous existence in distant brain structures may be a relevant and important point in the description of an integrative neurophysiology. As an aside, the comparative analysis of brain oscillatory responses in various brains (invertebrates, fish, cat brain, human brain) showed similar patterns of oscillatory responses to auditory stimulation, and to direct electrical stimulation in electroception and also in invertebrate ganglia [4]. Accordingly we assume that oscillatory behavior reflects basic network properties of the brain. Thus, for example, the functional meaning of the gamma band is probably wider

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

5J

100 msec after the sensory Hypotheses stimulation. The response has also a Neuroscientists discuss the “binding secondary oscillation with a latency problem”: How is the spatially dist :ibuted of approximately. 300 msec (see but temporally coherent (simulti neous) chapters 3 and 26 in [4]). electrical activity stemming from a large In a similar way, the 40 Hz response number of elementary neur:alcomponents is observed as a component of the vi- integrated to functional activity? (jamma sual EP in cats ([29] and see chapter rhythms have been suggestt:d to be associated with this binding phenomena: 3 in [4]). Gray and Singer [21-231 have reA similar phase-locked oscillation is ported that neurons in the cat visual also a component of the human audicortex exhibit oscillatory re ;ponses tory and visual response. in the frequency range of 40-60 Hz. The 40 Hz response that occurs 100 These oscillations occur - n synmsec following the applied stimulachrony with cells located viithin a tion is a pure sensitive component. functional column and are tightly Karakas and Basar [24] recently correlated with oscillatory f eld poshowed that, in the following paratentials. These researchers p -oposed digms, the 40 Hz response in the first that the synchronization of oscilla100 msec remained unchanged (i.e., tory responses of spatially distribit is not related to cognitive tasks): uted, feature-selective cells night be oddball (the subject is instructed to a way to establish relations between pay attention to a “deviant” stimulus features in different parts of’the viinterspersed into a sequence of “stansual field. dard‘’ stimuli); mismatch negativity Eckhorn, et al. [14-161, also found (MMN: the subject is concentrating stimulus-evoked resonances of on a task not related to the stimuli 35-85 Hz throughout the visual corwhile standard and deviant stimuli tex when primary coding channels are presented), and the “single stimuwere activated by their specific stimlus” paradigm after Polich. uli. They raised the question whether coherent oscillations do reflect a P300-40 Hz Component mechanism of feature linking in the A P300-40 Hz component has been revisual cortex. corded in the CA3 layer of the cat hippoDespite its current popularity, this campus by means of an ERP-paradigm hypothesis is probably insufficient to and using omitted stimuli following repetexplain the “ubiquity of gamma itive auditory stimulation as a target. A rhythms” [ 111. dominant and significant 40 Hz According to Llinis and Ribzry [27], time-locked wave packet occurs approxithe spatio-temporal magneticfield pattern mately 300 msec after the stimulation, but of gamma band activity suggests the presnot in the first 100 msec. ence of a coherent rostrocaudal sweep of activity repeating every 12.5msec due to a Attention-Related 40 Hz Activity continuous phase shift over th: hemiSeveral papers have noted the importance of sphere. The authors propose that this scan40 Hz activity in states of attention and moning occurs on account of thalamocortical tivation. Tiitinen, et al. [32], reported that resonant synaptic interactions. selective attention enhances the auditory 40 Hz transient response in humans, especially General View of Gamma Function. over the frontal and central areas. When interpreting gamma rhythms that Perceptual Switching may be universal building block:. instead A significant gamma-band (30-50 Hz) ac- of specific correlates of perceptt al bindtivity increase in the EEG during states of ing or higher cognitive processts, these “perceptual switching” was measured as results should be taken into acccunt. We induced by an ambiguous stimulus pat- propose that diffusely distributec gamma tern. The most significant 40 Hz enhance- generators are activated by different senments were measured in frontal areas and sory and cognitive events and react with can reach increases of 40 to 50% in states different time delays. This property of of naive and active observations. Obser- brain structures was called syncrzronized vation of ambiguous figures by naive sub- selectivities or synchroni.ied remnances jects induced 40 Hz wave packets in to point out that, upon stiniulatioii, generfrontal areas. ators of oscillatory potentials in the brain

.

.

. .

than the specific cognitive functions suggested by some authors.

Functions and Hypotheses Related to the Selectively Distributed Gamma Oscillations The functional correlates of the gamma oscillations described in the results presented in [3,4] include linking of percept u a1 inform at i o n : “binding ” [14-16, 21-23] and “vibrations in the memory” [13], leading to the question of whether “grandmother is an oscillation?’ [31]. Below we will deal with the diverse functional correlates of gamma activity, followed by a general view of gamma function. Olfuctory Bulb Freeman [ 181 and Freeman and Skarda [20] have shown that the EEG of the olfact o o bulb and cortex in awake and motivated rabbits and cats shows a characteristic temporal pattern consisting of bursts of 40-80 Hz oscillations, superimposed on a surface negative baseline potential shift coupled to each inspiration. Selectively Distributed Parallel-Processing Gumma System Since 1972 Basar’s laboratories have published approximately30 papers directly related to 40 Hz responses or indicating 40 Hz in complex response rhythms, finally proposing a working hypothesis that assumes the existence of a selectively distributed parallel-processing gamma system. Auditory gumma-band responses are selectively distributed in different brain structures (not only cortical but subcortical). The 40 Hz response is a robust, almost stable component of the auditory evokedpotential (EP) in cats. It can be recorded in the cortex, hippocampus, and cerebellum. It is phase locked and occurs in the first

.

58

.

.

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

Mar/June 1999

act in a coherent way. They may facilitate signal transmission by ways of resonance. The statements as listed above underline that gamma activity has multiple functional correlates. Alone, results presented in this report confirm that 40 Hz responses are involved with all types of cognitive and sensory processes, ranging from simple signal detection to complex CNS activity, such as recognition of ambiguous figures. The evaluation of the above statements also leads to the following generalization: 40 Hz signals of the brain (spontaneous, induced, or evoked) belong to important finctional building blocks of brain electrical activity. Also, 40 Hz wave packets may occur in different and distant structures, and may act parallel and separately depending on the modality of sensory or cognitive stimulation. They may show phase locking, time locking, or weak time locking.

Selectively Distributed Alpha Oscillations Properties and Functions Note that “alpha” is not noise-it is a quasi-deterministic signal (see chapter 10 in [ 3 ] ) . Alpha as Sensory Response in Human and Cat Brains. In chapters 2 and 10in [4] it was shown that inadequate stimuli could not generate significant and time-locked cortical alpha enhancements in the first 300 ms upon stimulation. The occipital cortex of the cat brain does not respond with enhanced 10 Hz if the stimulation is auditory (i.e., inadequate stimulation). Break of the Sensory Alpha Response Upon Injury to the Optic Nerve. Multiple sclerosis patients with opticus neuritis do not show alpha response after visual stimulation, thus clearly demonstrating the strong relation to sensory functions of the alpha response. According to the outlined working hypothesis indicating “a special transmission of sensory signals in the 10 Hz frequency range” one might suppose that the thalamus acts as a “gating structure” for the 10 Hz transmission (see chapter 27 in [4]). Long-Latency Alpha Oscillations. In these types of potentials in the poststimulus interval between 250 to 300 msec, a small 10 Hz enhancement can be recognized. The time locking is weak in comparison to results with adequate stimuli. Therefore, although single sweeps Moy/June 1999

may contain fairly high 10Hz amplitudes, the responses are not perfectly phase locked and, furthermore, they are delayed. These types of responses are also obtained in the auditory cortex by means of visual stimulation (inadequate stimulation, see chapters 2 and 10 in [4]).

Distributed Alpha Response to Sensory Stimuli in the Cat Brain. Since the days of Adrian, “evoked alpha” was a sign of the reactiveness of the CNS to sensory stimuli. Sensory-evoked 10 Hz responses were recorded in several cortical and intracortical structures (see chapters 1,9, 2, 10 in [4]). posed rhythm supplanted “the endogeAlpha oscillations with a duration of nous” one; or is it the same rhythm with a approx. 250 ms are recorded in the cerebel- shift in phase to carry the message; or is lum, reticular formation, and inferior the basic rhythm still there and another colliculus after auditory stimulation of 80 one added? In the latter case the brain dB and 2000 Hz. Alpha oscillations with a would have its own rhythm against which duration of approximately250 msec are re- to match this added one for both frecorded in the visual cortex, lateral quency and phase. geniculate nucleus, hippocampus, and reticular formationafter visual stimulation. Event-Related Synchronization and Desynchronization: Movement-ReProlonged Alpha Response Compolated Alpha. The terms “event-related nent of the P300 Response. Following desynchronization,” or ERD, and cognitive targets, the event-related alpha “event-related synchronization,” or ERS, oscillations are prolonged. During oddare used by Pfurtscheller and Klimesch ball experiments the response to the target [28] to describe the ability of neural struchas a duration up to 380 msec. This protures to generate more or less coherent oslonged oscillation contributes to the cillating potentials. ERD describes the 200/380 msec peaks. Our unpublished reattenuation or blocking, and ERS is the sults also show a prolonged 10 Hz reenhancement of oscillating potentials sponse to visual targets (see chapters 15, within the alpha and beta frequency 20, and 24 in [4]). bands. The Sensory Stimulationcan Evoke 10 Hz Oscillatory Behavior also at the Cellular Level [lo, 12,301. See chapter 9 in [3]. Alpha Can Be Induced. Perhaps the reader will find credible the claim that induced alpha-band rhythms are analogs of the gamma-band rhythms induced by moving stripes and odors in animal experiments. Consider closing the eyes in a lighted room: like turning off the lights, it reduces retinal illumination and initiates, or induces, an alpha-band burst in the EEG. Consider an example of alpha induced by (undefined) changes in internal state: the rhythm disappears the moment a subject follows the instruction “multiply 1 1 by 13,” and it reappears as soon as the answer is delivered. This finding of a rhythm phase-locked to stimulus shows that the brain now has a rhythm that has been imposed on it through a sensory system. Has this im-

Spontaneous and Evoked Alphas in Invertebrate Ganglia. Further evidence showing that the alpha activity cannot be explained through generators only in the thalamus or cortex are the recordings of the cerebral ganglion of Aplysia (see chapter 6 in [4]) and also with isolated ganglia of Helix pomatia. Electrically induced 10 Hz activity can be recorded in vitro in these small neural populations. Furthermore, amplitudes of spontaneous and induced alpha rhythms can be modulated by using pharmacological agents (see chapter 8 in [4]). Memory-Related, Event-Related 10 Hz Oscillations-Alpha Can Be Emitted. Basar, et al. [ 5 ] , demonstrated that a well-trained subject emitted time-locked bursts of alpha-band energy for up to a full second before the delivery of an expected target. In contrast to the modest evidence,

IEEE ENGINEERING IN MEDICINE AN0 BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

59

they advanced for anticipatory time-locking of waves in the gamma band that the alpha locking to the moment in the future when a target will be delivered is robust and highly significant statistically. Results of Klimesch [25] and of Basar, et al. [ 5 ] , (see chapter 22 in [4]) on dynamic memory demonstrate that alpha activity is strongly correlated with working memory and probably with engrams (traces) in long-term memory. According to Basar and coworkers, induced alpha rhythms can be considered to be internal EPs during periods of expectation of a visual or auditory target.

Hippocampal Event-Related Alpha Oscillations. According to the experimental results presented in chapters 1 and 27 in [4], all sensory stimulations evoke a lOHz oscillation with a duration of 250-300 msec in the hippocampus. The expression “every sensory stimulation” has to be pointed out here. The sensory stimulation may be an auditory one, which will evoke 10 Hz oscillatory waves, or it can be a visual one, which then triggers a 12 Hz oscillation immediately following this stimulation. This sensory information arises upon stimulation, most probably conveyed to the association cortex and to frontal nonlimbic associations cortex, as well as to other parts of the cortex, and also to the primary sensory areas. Hypotheses Alpha and Association Processes. The 10 Hz processes may facilitate, overall, association mechanisms in the brain. When a sensory or cognitive input elicits “10 Hz wave-trains’’ in several brain structures, then it can be expected that this general activity can serve as a communication signal “par excellence’’ between different structures (see chapter 24 in [4]).

Alpha Activity Controls EPs. As several results in chapter 8 and 12 in [3] also emphasize, the amplitude, time course, and frequency contents of EPs (especially N100-P200 wave complexes) strongly depend on the amplitude of alpha activity prior to a sensory stimulation. Selectively Distributed 10 Hz Oscillations. Rather than trying to locate a unique alpha generator, it is preferable to formulate the existence of a “selectively distributed parallel processing alpha system” (chapter 24 in [ 41). The physiological re60

sults concerning hippocampal alpha responses support this proposition: Auditory and visual stimulations elicit in cat hippocampus strong and stable alpha responses (10 Hz oscillations of approximately 300 ms), which are visible without filtering. Cortical and thalamic 10 Hz responses can be elicited only by stimulations that are adequate for the respective area. In contrast, hippocampal 10 Hz responses are present in all types of stimulations. In coherence functions computed from visually evoked responses, the hippocampo-cortical coherence is significantly larger than the thalamo-cortical coherence (the latter one being extremely low following inadequate stimulation). Thus, thalamo-cortical circuits are not unique in generating alpha responses; the hippocampus and formatio reticularis may even have a more general significance. In case the selectively distributed alpha system is not developed, no sensory 10 Hz response is registered: three-year-old children do not show any 10 Hz response either to auditory or to visual stimulation (“Brain response susceptibility”: see chapters 8, 13, and 14 in [3]).

.

.

.

.

Alpha as a Universal Code in the Brain. The statements above lead to the tentative interpretation of alpha as a universal code or universal operator in the brain. The major physiological meaning of 10 Hz oscillations is comparable to the putative universal role of gamma responses in brain signaling. Selectively Distributed Theta Oscillations Properties Theta Oscillations. Event-related theta oscillations are usually masked by alpha oscillations. Theta oscillations have larger amplitudes and more regular waveforms in frontal and central areas (Fz, Cz, Pz) (association areas). Theta oscillations in the cortex are most probably coupled with hippocampal theta generators during cognitive processes (see results and discussion in [9] and chapters 19 and 25 in [4]). Theta Frequency Components. Theta frequency components are either spontaneous patterns or induced oscillations. These components are manifested by the

utilization of “single trial theta epcchs” of EPs according to our combined analysis procedure. By describing a diffusely distributed parallel processing theta system on the entire brain, we emphasize that the theta generators must exist in seLera1 areas other than only in the llmbic SJ stem or frontal cortex. The existence of the marked theta responses in the visual cortex of the human brain is an example of the distributed nature of the parallel-processing theta systern (see chapters 9 and 10 in [4]).

Theta Activity Controls Frontal and Parietal Responses (see below for major operating oscillations). If pres timulus theta oscillations have low amplitudes, the NIOOP200 amplitudes of frontal EPs are increased, as demonstrated in [4]. Theta activity in the frontal part of the human brain controls the amplitude: of frontal EPs. In other words, the reaction of frontal areas of the human brain depend on its theta state. This selectivelq distrihuted system might have the strongest theta components in limbic structures or in cortical ones working strongly together with the limbic system (frontal cortex parietal cortex; see also chapter 19 in [4]). Functions Theta Response. Theta response is the most stable component of the tat P300 response. This response has its highest amplitude in the CA3 layer. According to Miller’s theory, cortico-hippocampal interaction and resonance play a basic role in cognitive processing (see chapter 25 in [4]).

Frontal Processing. Bimodal sensory stimulation induces great incr:ases in frontal theta response, thus demo istrating that complex events increase th,: frontal processing in the theta range. Event-Related Oscillations. ELvent-related oscillations in the theta band are prolonged andor have a second time window approximately 300 msec after the presentation of the target in oddball experiments. The prolongation of event-relaied theta oscillations contributes to the N2OO/P300 waveform (see chapters 19 and 20 in [4]). Prolongation of theta is irtterpret:d as being correlated with selective attention. Association. Event-related thet:. oscillations are also observed after an inadequate stimulation, whereas event-relatsd alpha

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

May/June 1999

oscillations are not existent if the stimulation is inadequate. Accordingly, the associative character for event-related theta oscillations is more pronounced than for higher-frequencyevent-relatedoscillations.

Multiple Functions of Major Operating Oscillations (Rhythms) in the Theta Band. Event-related potentials obtained with paradigms inducing focused attention, P300, and with stimuli giving rise to high expectancy states, have shown marked electrophysiological changes in the frontal cortex, parietal cortex, and limbic system. It was shown (chapters 20 and 21 in [4]) that the frontal areas of the human cortex reacted with enormous theta enhancements to cognitive stimulation requiring states of focused attention and short-term memory. In the human frontal cortex, a theta increase of 50% was recorded while a subject paid attention to a target that was expected (probability 100%). Similar experiments with cats demonstrated that also in the CA3 layer of the hippocampus, a theta increase of 40% was to be seen. In the P300 experiments, again learning tasks led to a theta increase with a time delay in frontal and parietal recordings. These results clearly demonstrate that cognitive tasks give rise to marked theta increases in EP components. When comparing the results of experiments with simple visual or sound stimulation, in which the EPs contain dominant alpha responses, we are inclined to state that cognitive loading increases the weight of theta components in comparison to alpha components. Furthermore, the increase in theta responses mostly takes place in frontal, hippocampal, or parietal structures. Even the omitted stimulus, which give rise to a P300 response in the cat’s hippocampus, has a dominant theta component, again with the largest component in the CA3 layer.

building blocks of functional signaling in the nervous system. However, as stated in chapter 29 in [4], a unique rhythm- or event-relatedoscillationcannot be the only functional processor for a given function. Complexjimctions are certainly intenuoven with several oscillations that may occur singly or in combination.

Carrier Signal for Cognition? If postulated that different types of projections might carry either theta or alpha information, the theta information would be probably transferred diffusely to several parts of the cortex as well as to association areas, which, in turn, may indirectly reach primary sensory areas of the cortex. This chain of events may even be developed so far as to assign to the theta responses a function as a carrier signal for cognitive-association processing. This working hypothesis does not only find support with the experiments described here, as the description of all experiments with association tasks indicated that theta enhancements occur regardless of the type of experiments and subjects used: Theta enhancements are very high in the hippocampus of the cat (chapters 18 and 16 in [4]) Theta enhancements in frontal and parietal recordings in experiments during states of focused attention and high expectancy (chapter 19 in [4]) Orientation. Table 25-1 in [4] shows are observed some examples of cat experiments during Miller’s results on cortico-hipexploration and search behavior, as well pocampal signal processing support as during motor behavior. Orienting is a the functional role of theta transmission in all cognitive states related to coordinated response that appears to indiassociation (for details, see chapter cate alertness, arousal, or readiness to 25 in [4]). . process information. However, the existence of a significant Hypotheses difference in the major operating oscillaComplex Functions. According to the tions in occipital or frontal areas gives statements above and the data presented in strong support to the possibility that theta several works (chapters 25,19, 20, and 15 (spontaneous, evoked, induced) and alpha in [4]), it is clear that event-relatedtheta os- rhythms (spontaneous, evoked, induced) cillations can be considered as important have fundamentally different functional

. .

-

Moy/June 1999

r

Gate Thalamus. Seemingly, if a sensory stimulation already in the gate thalamus in the 10 Hz frequency channel is interrupted, the theta component of the same sensory message is not blocked in the thalamus but is conveyed further into the cortex. This working hypothesis can be better understood or can find more support from analysis of electrophysiological behavior in the hippocampus using single EEG-EP epochs.

operations. But, during some functional states, major operating rhythms can change their functional roles; the nature of the experiment, i.e., the task, can influence the weight of these functional components on brain rhythms. Selectively Distributed Delta Oscillations Sensory Function The delta-responseoscillations(the delta response) are recorded in auditory and visual EPs in all scalp recordings of the human brain and also in cortical and intracortical structures of the cat brain. This means that sensations evoke a delta response; i.e., delta-response oscillations are encountered as sensory components in brain responses (chapters 1, 2,9, and 10 in [4]).

Cognitive Functions The amplitude of the delta response is considerably increased during oddball experiments. The delta response to the target signals are delayed and prolonged. Accordingly, it was concluded that the delta response is related to signal detection and decision making (see chapters 20 and 21 in [4]). Cognitive Delta Response is Selectively Distributed in Humans The responses to visual targets have their greatest response amplitudes in parietal locations, whereas following auditory target signals, the greatest delta response amplitudes are observed in central and frontal areas (see chapters 20,18, and 28 in [4]). Signal Detection at Hearing Threshold Human subjects do respond with delta oscillations at the hearing threshold. Ac-

IEEE ENGINEERING IN MEDICINE AN0 BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

61

We will now summarize principles and theories that emerged from reports and results of conferences that were published from our research group [3,4]. To begin, we will treat an important question posed by Fessard almost 40 years ago. Basar and Karakas (in [4]) outlined a satisfactory approach to the main claim of Fessard: an integrative well-functioning system must obey some general rules or laws, even if it is a very complex one.

Oscillations Govern the General Transfer Functions in Neural Tissues of the Brain Fessard [17] emphasized the role of neuronal networks in the brain as follows:

cordingly, the role of the delta response in signal detection and decision making is demonstrated in different types of cognitive experiments (see chapter 12 in [4]).

Mismatch Negativity The mismatch negativity (MMN) is shaped by a delayed delta response superimposed with a significant theta response. Sleep Delta Responses Sleep delta responses are probably the major processing signals in the sleepingbrain.

Multifold Functions We have described approximately 50 types of oscillatory activities, with definitive or tentative explanations of their functional relations (there are more examples in [3,4]; however, we do not summarize all of them here). The results presented clearly demonstrated that it is not possible to assign to a given type of oscillatory activity to only one function. These oscillations have multifold functions and act as universal operators or codes of brain functional activity. Moreover, besides the frequency and site of the activity, some other parameters are also involved (or interwoven) in brain functioning. These parameters are enhancement, time locking, phase locking, delay of the oscillation, and prolongation of oscillations. Complex and integrative brain functions are manifested in the superposition of several oscillations and in frequency stabilization, and also in the following parameters: degree of prolongation, enhancement, delay, time-locking, and phase-locking in several time windows.

The brain, even when studied from the restricted point of view of sensory communications, must not be considered simply as a juxtaposition of private lines, leading to a mosaic of independent cortical territories, one for each sense modality, with internal subdivisions corresponding to topical differentiations. The track of a single-unit message is doomed to be rapidly lost when one tries to follow it through a neuronal field endowed with network properties, within which the elementary message readily interacts with many others. Unfortunately, we still lack principles that would help us describe and master such operations in which heterosensory communications are involved. These principles may gradually emerge in thefuturefrom an extensive use of multiple microelectrode recordings, together with a systematic treatment of data by modern electronic computers, so that pattern-to-pattern transformation matrices can be established and possibly generalized. For the time being, it seems that we should do better to try to clear up such principles as seem to govern the most

general transformations-or

transfer

functions-of multiunit homogeneous messages during their progressions through neuronal networks. The transfer function measures the ability of a network (here, neural networks of the brain) to increase (facilitate) or impede (inhibit) transmission of signals in given frequency channels. Thus, the properties of the transferfinction constitute the main framework for signalprocessing and communication. The existence of general transfer functions would then be interpreted as the existence of networks distributed in the brain that

show similar frequency characteri:,tics or facilitate or even increase the signal transmission in given frequency chanrels. In an electric system, optimal transmission of signals is often reached when distributed subsystems of the system are timed to the same frequency range. lloes the brain have such subsystems tuned. in similar frequency ranges, or do there exist common frequency modes in the brain? 3esults discussed in [3, 41 provide a real, definitive approach to this question. The described frequency characteristics in all brain tissues that embrace the resonant oscillatory processes or selectively distributed oscillatory systems of thr: brain (delta, theta, alpha, beta, gamma) constitute and govern mathematically t le general transfer functions of the brain [see the definition of frequency characteristics in chapter 4 in [3]). All brain tissues, both in the animal and human brain (including isolated ganglia of invertebrates, of low verkbrates, and of the human brain) react to sensitive and cognitive inputs with oscillalory activity within almost invariant and ,penera1 governing frequency channels. Experimental results show that the degree of synchrony, amplitude, duration, ant1 phase lag continuously vary, but similar oscillations are always present in the activated brain tissues. Synopsis of the Theory of

Functional Brain Oscillatilms Spontaneous Oscillations Natural Frequencies of the Brain. The brain has several types of natural oscillatory activities in various fre’juency ranges; delta, theta, alpha, gamma, and high frequencies. These oscillaticas may occur spontaneously and can be zmitted, induced, or evoked. The EEG Is Not Noise: Chaotic !signals. The natural oscillations are not always noise but probably reflect properties of chaotic attractors. We use the expression “quasi-deterministic” to describe this oscillatory behavior. Natural Frequencies (EEG frequencies) Are Recorded at the C’ellular Level. Spontaneous and evoked or induced rhythms are also observed i i 10 Hz, 40 Hz, beta, and delta freyuencii s at the cellular level. The descriptions in chapter 9 in [3] clearly show that one can correlate the spikes at the cellular level with field potentials (EEG-EPs). Although the rela\

62

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

M a y h e 1999

tion between EEG and single cell activity was also demonstrated earlier, new results of Llinfis [26]; Singer, et al. [21-231; Eckhorn, et al. [14-161; and Dinse, et al. [ 121; by using multiple electrodes extended this approach, now almost reaching perfection. Together with the nonlinear approach, electroencephalography will probably gain importance for describing further the states of the brain.

Major Operating Rhythms. Experimental results presented in [3,4] demonstrate that the EPs and the EEG immediately prior to stimulation build an entity with the next EP. The results described in chapter 12 in [3] further demonstrate that these are major operating rhythms in several areas of the cortex. These major operating rhythms also influence the shape of each potential in different frequencies. A good example is that the theta activity is the major operating rhythm in the frontal cortex. On the contrary, the major operating rhythm of the occipital cortex is 12 Hz oscillation. Event-Reluted Oscillations Properties of the Transfer Functions. The properties of the transfer functions interwoven with resonances and natural frequencies constitute the main framework for signal processing and communication: alpha, theta, gamma, delta responses provide the main electrical information processing in the brain.

Transitions of EEG from Disordered to Ordered States. The frequency-domain description of EPs in the cortex, thalamus, reticular formation, hippocampus, and cerebellum show a similar overall frequency structure to the ongoing activity (EEG), indicating a resonating universal mechanism: The sensory stimulus brings the brain into a more coherent state. In response to the stimulus, the frequency bands in various structures of the brain become much sharper and narrower, and they become coherent in phase and frequency. In a given frequency channel, the magnitude of the response is enhanced against the magnitude of the ongoing activity (see chapter 1 1 in [ 3 ] ) . Response Susceptibility of the Brain. The results led us to derive a “rule of excitability” that can be stated as follows. If a brain structure shows spontaneous rhythmic activity in a given frequency channel, then this structure is also in the Moy/June 1999

same frequency channel, and, moreover,

it will produce internal EPs in response to internal afferent impulses originating in the CNS, or respond in the form of EPs to external sensory stimuli, with patterns similar to those of internal EPs (see chapter 14 in [3]).

Superposition Principle. EPs are a superposition of delta, theta, alpha, and gamma oscillations that are enhanced or phase-locked depending on the nature of stimuli (sensory or cognitive). Brain Real Responses. By using several physiological, psychological, or biophysical methods and strategies, it was shown that the evoked rhythms ranging from delta to gamma frequencies are real brain responses that are related to functions (see chapter 18 in [3]). Internal Evoked Potentials. The findings on stereodynamics of brain potentials show that the evoked responses in all the nuclei and in all the frequencies are strongly dependent on the spontaneous activities just prior to stimulus. There are cases in which the filtered EEG-EP-epochs already depict, in the EEG portion, ample potentials similar to the filtered EP signals taking place immediately after stimulation. The resemblance in the shapes of EPs and such EEG bursts leads us to use the expression of “internal evoked potentials” for the description of large amplitude and synchronized EEG recordings. When the relevant internal EPs are recorded before the stimulation, usually the EPs induced by stimulation do not have large amplitudes. Accordingly, we assume that EP research with single EPs will be useful also in contributing to the understanding of EEG population dynamics. Functional Meaning Multiple or Diverse Functions Related to Oscillations. On one hand, spontaneous and event-related oscillations are correlated with several brain functions. On the other hand, functions are related to a superposition of oscillatory responses. The amplitude and duration of the oscillation are also correlated with function and are modified according to functional changes. EP components which we called “alpha-”, “theta-”, “delta-” “beta-” and “gamma responses” are correlated to various functions as demonstrated in [3,4] and summarized in chapter 30 in [4]. Superpo-

sition of all these oscillatory signals can be related to complex functions. The complexity of an event is not reflected only by EPs. In a number of experiments, the spontaneous EEG can be also considered as internally evoked potentials where the inputs are coming from yet hidden sources.

EEG as a Functional Brain Code, Oscillations as Brain Alphabet. The core concept of the “EEG-Brain Dynamics” conference (as summarized in [ 3 ] ) was based on the statement that the EEG is not simply a noise, but it is, in all probability, one of the most useful signals of the brain related to EPs. This core concept has been fortified and extended to include brain function related to EEG (see [3, 41). The EEG itself is considered here not only as a quasi-deterministic signal, but also as the most useful activity in its capacity as brain code for function. EEG Frequency Generators are Selectively Distributed. According to our hypothesis explained in [3,4], all the EEG frequency generators are selectively distributed within the entire brain. Sensory or cognitive inputs bring these EEG generators into a resonating state. The described selectively distributed oscillatory systems govern the excitability and communication of all brain structures. Not only in the cortex but also in deeper structures of the brain (for example, in the brainstem), 10 Hz, 4 Hz and other resonances can be registered. During such activity, the brain reaches coherent states with regard to time and space. Not only the cortex, but the entire brain is involved in complex re-

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

63

sponses, such as the P300 response. In some of the functional reactions, only a few EEG components are involved, and one can sometimes encounter simple oscillatory responses; for example, the delta response at the hearing threshold. Other typical examples are experiments during which the subjects pay attention to repetitively applied signals, some of which were omitted. In this case, the frontal event-related potentials elicited by signals preceding the omitted stimulus are reduced almost to a unique and homogeneous theta oscillation. It is further proposed that event-related oscillations behave like “letters of the alphabet” and complete brain functions are reflected as “words” constituted of letters such as alpha, beta, gamma. Selectively distributed oscillatory networks in frequency bands of delta, theta, alpha, beta, and gamma play a major role within the brain and for sensory and cognitive brain functioning. Not only the “type” but the “size” and “combinations” of the letters describe functions (see chapter 29 in [4]). Sensations and cognitive events evoke superimposed oscillations that are transmitted to all brain tissues with various degrees of intensity, synchronization, duration, and delay almost in parallel. Depending on the sensory and cognitive nature of the inputs and side of recordings, oscillations in these frequency channels are following the input. The various types of oscillations contributing to or manifesting brain functioning are described in chapter 30 in [4].

Real Advantages. Hippocampal alpha networks also control the alpha response. The widely held theory on alpha generators is based on the existence of 64

thalamo-cortical circuits (see chapter 24 in [4]). The results presented in [3, 41, however, demonstrate the existence of hippocampal alpha generators that control the alpha enhancements to all types of sensory modalities. These results have been achieved through cross-modality experiments in the cat brain as well as through the utilization of the coherence function (see chapter 1 in [4]). These results contradict the mainstream theories for alpha generators such as the facultative pacemaker theory described in chapter 24 in [4]. It is important to point out that only an integrative analysis using data from the entire brain can enable the establishment of theories regarding generators. Such an analysis can be made only by using semi-microelectrodes, as was the case with our analysis, and not by single-unit studies.

Integrative Brain Activity is Manifested in EEG and Event-Related Oscillations. The examples stated in [3,4] for the description of cognitive potentials in the cat or human brain were obtained under specific experimental paradigms. The functional activities of the CNS are certainly not limited to the experiments and results that are covered in [3,4]. The core concept is related to the utilization of the EEG code as an important key for understanding the integrative activity of the brain.

A “Neurons-Brain” Doctrine: New Thoughts There are several trends to extend or renew the neuron doctrine of Sherrington. Freeman proposed a ‘hew Sherringtonian view” of integrative brain function, whereas Barlow and Mountcastle proposed modern views on the neuron doctrine (see chapter 28 in [4]). Relying on the above outlined results and on three volumes recently published [3, 4, 71, we present here a new doctrine to describe a basic framework for signaling of communication and functional operation in the brain. This new doctrine cannot be perfect and cannot cover all results and principles accumulated in the recent decades. However, it provides a plausible and progressive framework, which should replace the old doctrine. The latter should be considered as a special case of the more general new “neurons doctrine.” 1. The neuron is the basic signaling element of the brain.

2. Oscillatoryneural acrivity is :onsidered as a basic signal reflecting natural frequencies of the brain (this thesis relies on works of Verzeano [33]; Gray and Singer [23]; Silva, et al. [30]; EcEhorn et al. [ 14-16];Dinse, et al. [ 121; see chapter9 in [3]). 3. Neural assemblies replace the neuron in the description of complex brain functions. This view diverges from Sherrington’s “neuron doctrine.” As a metaphor to physical sciences, neurons can be considered as atoms-neural assemblies that contribute to a function, as molecules. Accordingly, the metiphor is similar to statistical mechunics andor gas laws (see chapter 4.2 in [3]); a landmark is the important work by Freeman [ IS]. 4. Oscillatory activities (event.related, induced, or spontaneous) govern .he most general transfer functions in the brain (frequency characteristics anc power spectra are governed with alpha, gamma, theta, delta, etc., oscillations-which is confirmed by the wavelet approach). Further, as stated in chapter 3 l in [4], the general t r a n s f e r f u n c t i o n s p r o v i d e a framework for electrical informa:ion processing in the brain. 5. Oscillations in different frequency ranges are a property of the neurons (see No. 2 above). Selectively distributed oscillatory neural populations, however, behave with “molecular properties” (alpha, beta, gamma, delta, theta, being “atomar properties”). These oscillatory networks are activated upon sensory stimulation or event-related tasks by nanifestation of synchronization of neural activity; partial synchrony; enhancements; or blocking or desynchrclnization of oscillations depending on the nature of the sensation or event, and, accordingly, depending on the function pe formed. These selectively distributed networks are operators of general brain functions including communication and as:.ociation and data retrieval (molecular properties, see previous item). 6. Major operating rhythms play a key role in association and communication. Topological distribution of oscillators is heterogeneous and their functions are multifold. Accordingly, parallel processing is not perfect between distributed pol~ulations since the major operating rhythnis are selectively distributed. For examFle, alpha responses to light stimulation do not appear in the medial geniculate nucleus and in the auditory cortex, whlereas tlie lateral

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

Mty/June 1999

geniculate nucleus and the primary visual cortex respond with large alpha enhancements (more examples are summarized in chapter 30 in [4]). 7. Types ofneurons do not play a major role for frequency tuning of oscillatory networks. The neural architectonics of the cerebellar cortex, cerebellum, and hippocampus are completely different. In spite of this, all these structures behave with almost similar frequency responses. 8. Distributed oscillatory networks react selectively upon application of pharmacological agents. For example, the neuropeptide ceruletide causes a great change in hippocampal evoked response, which takes the form of a homogeneous 3-4 Hz response, whereas the frequency response of the cerebellum remained completely unchanged since this agent does not have any action in the cerebellum. Acetylcholine enormously activates the 4 Hz response of the hippocampus, whereas other structures are less influenced (see chapter 8 in [4]). 9. Functions in the brain are manifested by varied degrees of superpositions of oscillations in EEG frequency ranges. There are varied degrees of responsiveness depending on the strength of the stimulation or the event presented to the CNS. Accordingly, neuron assemblies do not react with all-or none behavior, as in the single neuron doctrine. 10. There exists a strong inverse relation between prestimulus oscillations and brain responses. Spontaneous oscillations control the amplitude and shape of population responses (see chapters 12 and 14in PI and 121).

Pro$ Dr. Erol Basar is professor of biophysics and physiology and is at present the head of the Neurophysiology Research Group at the Medical University of Lubeck. He is also honorary president of the Brain Dynamics Unit of the Turkish Research Council in Ankara. He studied at the universities of Munich and Hamburg and was physiology-instructor from 1965-68 at Hamburg University. Dr. Basar was awarded his Ph.D. degree at the University of Hannover in 1968. Following a period of research at the Rockland Brain Research Center in New York, he worked as a professor at the University of Hacettepe in Ankara between 1970 and 1978. He was then appointed as the R. Merton visiting professor at the University of Kiel, after which he took up an appointment at the Medical University of May/June 1999

Liibeck in 1980. The current research of Dr. Basar covers EEG, event-related potentials, and chaos integrative aspects of neurophysiology in a broad sphere of experiments, ranging from invertebrates to cats and in the human brain. He is also teaching physiology to medical students.

Pro$ Dr. C. Basar-Eroglu was born in 1950 in Ankara, Turkey, where she received her M.Sc. degree in biology and Ph.D. degree in biophysics at the Hacettepe University of Ankara. In 1992 she was awarded the “venia legendi” at the Medical University of Liibeck. Appointed professor in 1997, Dr. Basar-Eroglu is a psychophysiologist at the Institute of Psychology and Cognition Research, University of Bremen, Germany. Her current research activities focus on the electroencephalogram, event-related potentials, cognitive processes of the brain, and multistable perception. Pro5 Dr. Sirel Karakas was born in Zonguldak, Turkey, in 1942. She graduated from Middle East Technical University in Turkey (1963, received her M.S. degree in experimental psychology (1971), and her Ph.D. in b i o p h y s i c s (1978). She is presently working at the Hacettepe University as Professor of Experimental Psychology. Professor Karakas has established a Cognitive Psychophysiology Research Center at the Experimental Psychology Department at Hacettepe University and has acted as codirector at the Brain Dynamics Research Unit (1993-1998) of the Scientific and Technical Research Council (TUBITAK) of Turkey. She is a member of the Brain Dynamics Research Network of TUBITAK. Her research interests are on cognitive psychophysiology (eventrelated potentials) and neuropsychology with theoretical interest in the development of integrative models for human neurocognition. Dr. Martin Schurmann was born in Bottrop, Germany, in 1962. He received the degree “Dr. med.” (M.D.) from the Rheinisch-Westfalische Technische Hochschule Aachen, Germany, in 1988. From 1988 to I989 he worked in the Department of Neuroanatomy at the Rheinisch-Westfalische Technische Hochschule in Aachen, Germany. In 1989 he joined the neurophysiology group at the Institute of Physiology,

Medizinische Universitat zu Lubeck, Germany, where he received the degree “Dr. med. habil.” (Ph.D.) in 1997. He is currently working in the field of EEG and evoked-potential analysis with respect to functional roles of event-related EEG oscillations in sensory and cognitive processing.

Address for Correspondence: Dr. Erol Basar, Institut fur Physiologie, Medizinische Universitat zu Liiheck, 23538 Liibeck, Germany. Fax: +49 451 500-4171. E-mail ebasar@physio. mu-1uebeck.de

References I . Abraham RH, Shaw CD: Dynamics: The Geometry of Behavior, Vol. 3. Aerial, Santa Cruz, CA, 1983. 2.Arieli A, Sterkin S, Grinvald A, Aertsen A: Dynamics of ongoing activity: explanation of the large variability in evoked cortical responses. Science 272: 1868-1871,1996. 3. Basar E: Brain Function and Oscillations. I. Brain Oscillations. Principles and Approaches. Springer: Berlin, Heidelberg, New York, 1998. 4.Basar E: Brain Function und Oscillations. II. Integrative Brain Function. Neurophysiology and Cognitive Processes. Springer: Berlin, Heidelberg, New York 1998). 5. Basar E,Basar-Eroglu C, Roschke J, Schiitt A: The EEG is a quasi-deterministic signal anticipating sensory-cognitive tasks. In Basar E., Bullock T.H. (eds),Brain Dynamics. Springer: Berlin, Heidelberg, New York, pp. 43-71,1989. 6. Basar E, Bullock TH (eds.): Induced Rhythms in the Bruin. Birkhauser: Boston, Basel, Berlin,

1992. 7. Basar E, Hari R, Lopes da Silva FH, Schiirmann M (eds.): Brain alpha activity - New aspects and functional correlates. Int J Psychophysiol26: 1-482,1997. 8.Basar E, Rahn E, Demiralp T, Schiirmann M: Spontaneous EEG theta activity controls frontal visual evoked potential amplitudes. Electroencephalogr. Clin Neurophysiol 108: 101-109,1998. 9. Basar-Eroglu C, Basar E, Demiralp T, Schiirmann M: P300-response: Possible psychophysiological correlates in delta and theta frequency channels. A review. I n t J Psychophy.Yio1, 13: 161-179 1992. IO.Creutzfeld OD,Watanabe S,Lux HD: Relations between EEG-phenomena and potentials of single cortical cells. I. Evoked responses after thalamic and epicortical stimulation. Electroencephalogr Clin Neurophysiol. 20: 1-1 8, 1966. 1 I. Desmedt JE, Tomherg C: Transient phase-locking of 40 Hz electrical oscillations in prefrontal and parietal human cortex reflects the process of conscious somatic perception. NeurosciLett, 168: 126-129, 1994. 12. Dinse HR, Kriiger K, Akhavan AC, Spengler F, Schoner G, Schreiner CE:

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

65

Low-frequency oscillations of visual, auditory, and somatosensory cortical neurons evoked by sensory stimulation. Inr J Psychophysiol 26: 205-227, 1997. 13. Douglas RJ,Martin KAC: Vibrations in the memory. Nature 373: 563-564, 1995. 14. Eckhorn R, Bauer R, Jordan W, Brosch M, Kruse W, Munk M, Reithoeck HJ: Coherent oscillations: a mechanism of feature linking in the visual cortex? Biol Cybern 60: 121-130, 1988. 15. Eckhorn R, Bauer R, Reitboeck HJ: Discontinuities in visual cortex and possible functional implications: relating cortical structure and function with multielectrode/correlation techniques. In: Basar, E., Bullock, T.H. (eds.), Brain Dynamics - Progress and Perspectives. Springer: Berlin, Heidelberg, New York, pp. 267-278 (Springer Series in Brain Dynamics, Vol. 2), 1989. 16. Eckhorn R, Reithoeck HJ, Arndt M, Dicke P: Feature linking via stimulus-evoked oscillations: experimental results from cat visual cortex and functional implications from a network model. Proc Confon Neural Nehvorks, Washington, DC, 1989. 17. Fessard A: The role of neuronal networks in sensory communications within the brain. In Rosenblith W.A. (ed.), Sensoty Communication. MIT Press, Cambridge, MA, 1961. 18. Freeman WJ: Mass Action in the Nervous Svstem. Academic Press, New York, 1975.

19. Freeman WJ: Foreword. In: Basar E., Brain Function and Oscillations, vol.I and 11. Springer: Berlin, Heidelberg New York, 1998. 20. Freeman W J , Skarda C A : Spatial EEG-patterns, non-linear dynamics and perception: The neo-Sheningtonian view. Brain Res Rev IO: 147.175, 1985. 2 I. Gray CM, Konig P, Engel AK, Singer W: Oscillatory responses in cat visual cortex exhibit inter-columnar synchronization which reflects global stimulus properties. Nature 338: 334- 337, 1989. 22. Gray CM, Singer W: Stimulus-specific neuronal oscillations in the cat visual cortex: a cortical function unit. Soc Neurosci Abstr 404: 3, 1987. 23. Gray CM, Singer W: Stimulus-specific neuronal oscillations in orientation columns of cat visual cortex. Proc Natl Acad Sri (USA). 86: 1698.1702, 1989. 24. Karakas S, Basar E: Early gamma response is sensory in origin: A conclusion based on cross-comparison of results from multiple experimental paradigms. Znt J Psychophysiol, in press. 25. Klimesch W: Memory processes, brain oscillations and EEG synchronization. Int J Psychophysiol24: 61-100, 1996. 26. Llinas R: The intrinsic electrophysiological properties of mammalian neurons: Insights into

central nervous system function. Science 242: 16.54-1664, 1988. 27. Llinas RR, Ribary U: Rostrocaudal scan in human brain: A global characteristic of tEe 40-Ha response during sensory input. In Basar Z . , Bullock T.H. (eds.), Induced Rhythms in the Brain. Birkhauser, Boston, pp. 147-154, 1992. 28. Pfurtscheller G, Klimesch W: Event-related synchronization and desynchronization of alpha and beta waves in a cognitive task. In Ilasar E., Bullock, T.H. (eds.), Induced Rhythms in the Brain. Birkhauser, Boston, pp. I 17.128, 1992. 29. Schiirmann M, Basar-Eroglu C, llasar E: Gamma responses in the EEG: elementary signals with multiple functional correlates. Neu -oReport 8: 1793-1796, (1997). 30. Silva LR, Amitai Y, Connors BW: Intrinsic oscillations of neocortex generated by layer 5 pyramidal neurons. Science 2.5 1 : 432-435. 199I . 3 1 . Stryker MP: Is grandmother an oscillation? Nature 338: 297-298, 1989. 32. Tiitinen H, Sinkkonen J, Reinikainen K, Alko K, Lavikainen J, Naatanen R: Selective attention enhances the auditory 40-Hz transient response in humans. Nature 364: 59-60. 1993. 33. Verzeano M: The study of neuronal ietworks in the mammalian brain. In Thompson R.F., Patterson M.M. (eds), Bioelectric R.icordirig Techniques.PartA. Cellular Processes c nd Bruin Potentials. Academic Press, New York, 1973.

IEEE Transactions on Rehabilitation Engineering1 Search for Editor he IEEE EMBS is currently undertaking a search for Editor of the Transactions on Rehabilitation Engineering (TRE). This position is for a threeyear term and is renewable for an additional three years. The term of appointment would preferably start by January, 2001, and an overlap period of 3-6 months with the existing editors during 2000 will be supported to facilitate transfer of responsibilities. At the time of appointment, the successful applicant must be a member of the EMBS. The IEEE Transactions on Rehabilitation Engineering focuses on the rehabilitation engineering aspects of biomedical engineering and covers such topics as: biomechanics and analysis of human movement human performance measurement and analysis nerve stimulation control and analysis of prosthetics and orthotics signal processing for rehabilitation applications fundamentals and innovationsin assistive technology computer software and hardware for rehabilitation

66

This journal is supported by the IEEE EMBS and co-sponsored by RESNA. Interested individuals should communicate electronically or by post, outlining the following: their goals for the Transactions a brief synopsis of their expertise in technical editing and rehabilitation engineering a curriculum vitae Preference will be given to candidates with a clear interest andor experience in the use of web-based paper submission and review. Applications will be accepted until a suitable candidate is selected. Review of applications by the search committee will commence in October 1999. Auulicatlons should be sen1 10: IEEE EMBS Secretariat, c/o National Research Council of Canada Building M-19, Room 220 Ottawa, ON Canada KIA OR6 FAX: (613) 954-2216 E-MAIL: [email protected]

IEEE ENGINEERING IN MEDICINE AND BIOLOGY

Authorized licensed use limited to: ULAKBIM UASL - ISTANBUL KULTUR UNIVERSITESI. Downloaded on November 3, 2008 at 03:50 from IEEE Xplore. Restrictions apply.

Moy/June 1999