Annual Reviews OnlineSearch Annual ReviewsAnnual Review of Neuroscience OnlineMost Downloaded NeuroscienceReviewsMost Cited NeuroscienceReviewsAnnual Review of Neuroscience ErrataView Current Editorial CommitteeAll Articles in the Annual Review of Neuroscience, Vol. 32 Neuropathic Pain: A Maladaptive Response of the NervousSystem to DamageSynaptic Mechanisms for Plasticity in NeocortexNeurocognitive Mechanisms in Depression: Implications forTreatmentUsing Diffusion Imaging to Study Human Connectional AnatomySerotonin in Affective ControlPhysiology and Pharmacology of Striatal NeuronsThe Glial Nature of Embryonic and Adult Neural Stem CellsRepresentation of Number in the BrainNeuronal Gamma-Band Synchronization as a Fundamental Processin Cortical ComputationThe Neurobiology of Individual Differences in ComplexBehavioral TraitsThe Science of Neural Interface SystemsThe Neuropsychopharmacology of Fronto-Executive Function:Monoaminergic ModulationThe Influence of Stress Hormones on Fear CircuitryThe Primate Cortical Auditory System and Neural Representation ofConspecific VocalizationsEstablishment of Axon-Dendrite Polarity in Developing NeuronsAxon Growth and Guidance: Receptor Regulationand Signal TransductionCerebellum and Nonmotor FunctionAdvances in Light Microscopy for NeuroscienceANRV379-NE32-09 ARI 10 May 2009 8:58Neuronal Gamma-BandSynchronization as aFundamental Processin Cortical ComputationPascal FriesDonders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen,Nijmegen, The Netherlands; email: [email protected]. Rev. Neurosci. 2009. 32:209–24First published online as a Review in Advance onMarch 20, 2009The Annual Review of Neuroscience is online atneuro.annualreviews.orgThis article’s doi:10.1146/annurev.neuro.051508.135603Copyrightc 2009 by Annual Reviews.All rights reserved0147-006X/09/0721-0209$20.00Key Wordsrhythm, oscillation, theta, alpha, attention, communicationAbstractNeuronal gamma-band synchronization is found in many cortical areas,is induced by different stimuli or tasks, and is related to several cognitivecapacities. Thus, it appears as if many different gamma-band synchro-nization phenomena subserve many different functions. I argue thatgamma-band synchronization is a fundamental process that subservesan elemental operation of cortical computation. Cortical computationunfolds in the interplay between neuronal dynamics and structural neu-ronal connectivity. A core motif of neuronal connectivity is conver-gence, which brings about both selectivity and invariance of neuronalresponses. However, those core functions can be achieved simultane-ously only if converging neuronal inputs are functionally segmentedand if only one segment is selected at a time. This segmentation andselection can be elegantly achieved if structural connectivity interactswith neuronal synchronization. I propose that this process is at least oneof the fundamental functions of gamma-band synchronization, whichthen subserves numerous higher cognitive functions.209Annu. Rev. Neurosci. 2009.32:209-224. Downloaded from www.annualreviews.orgby University of Arizona Library on 01/23/11. For personal use only.Click here for quick links to Annual Reviews content online, including:• Other articles in this volume• Top cited articles• Top downloaded articles• Our comprehensive searchFurtherANNUALREVIEWSANRV379-NE32-09 ARI 10 May 2009 8:58ContentsINTRODUCTION .................. 210THE SCOPE OF NEURONALGAMMA-BANDSYNCHRONIZATION ........... 211THE MECHANISMS BEHINDGAMMA-BANDSYNCHRONIZATION ........... 212IMMEDIATE CONSEQUENCESOF GAMMA-BANDSYNCHRONIZATION ........... 212Feedforward CoincidenceDetection ...................... 213Rhythmic Input Gain Modulation . . 213NETWORK CONSEQUENCES .... 215BIASED COMPETITION ........... 216SELECTIVITY AND INVARIANCE,SEGMENTATION, ANDSELECTION ..................... 217EXPLORATION THROUGHRHYTHMIC BREAKINGOF GAMMA ...................... 217CONCLUSION ..................... 220INTRODUCTIONGamma-band synchronization took centerstage in systems neuroscience in the late 1980sand early 1990s, primarily through the workof Singer and colleagues (Eckhorn et al. 1988,Gray et al. 1989, Engel et al. 1991b, Singer& Gray 1995). These pioneering studies fo-cused on testing the proposal that gamma-bandsynchronization subserves perceptual binding,often referred to in short as the binding by syn-chronization hypothesis. The numerous stud-ies supporting this proposal as well as the fewstudies that presented apparently contradict-ing evidence have been reviewed extensivelyover the years (Singer 1999). I therefore focuson ideas about gamma-band synchronizationthat are in accordance with the binding-by-synchronization hypothesis but take a differentperspective.The binding-by-synchronization hypoth-esis is a psychophysiological hypothesis,proposing a physiological observable as the ma-terial substrate of a psychological experience.In a conventional psychophysiological hypoth-esis, enhanced activity of visual cortical neu-rons might be proposed as the substrate ofthe perception of the inducing visual stimu-lus. The binding-by-synchronization hypoth-esis applies an ingenious modification of theconventional approach by proposing that re-lations between physiological observables arethe substrate of the experience of relationsamong stimuli. Specifically, the synchroniza-tion among neuronal activities was proposed asa correlate of perceptual binding between thestimuli inducing the synchronized activities.The approach that I present here might becalled physiophysiological, because it primar-ily considers the relation between two physio-logical phenomena, namely neuronal synchro-nization and neuronal interactions. I begin byconsidering the widespread, yet specific, occur-rence of the phenomenon of neuronal gamma-band synchronization. This widespread naturebegs the question of whether gamma-band syn-chronization affects neuronal processing. Thisquestion is of central importance: If gamma-band synchronization has no consequences forneuronal processing, then it is an epiphe-nomenon, but if it has consequences, then ithas some function whatever the precise func-tion may be. To decide between the latter twoalternatives, I quickly review the mechanismsthat underlie gamma-band synchronizationbecause they suggest potential immediate con-sequences for neuronal processing. I then re-view the actual evidence for such immediateconsequences and proceed to