Restricted access

Abstract

This article concerns the nature of evoked brain responses and the principles underlying their generation. We start with the premise that the sensory brain has evolved to represent or infer the causes of changes in its sensory inputs. The problem of inference is well formulated in statistical terms. The statistical fundaments of inference may therefore afford important constraints on neuronal implementation. By formulating the original ideas of Helmholtz on perception, in terms of modern-day statistical theories, one arrives at a model of perceptual inference and learning that can explain a remarkable range of neurobiological facts.

It turns out that the problems of inferring the causes of sensory input (perceptual inference) and learning the relationship between input and cause (perceptual learning) can be resolved using exactly the same principle. Specifically, both inference and learning rest on minimizing the brain's free energy, as defined in statistical physics. Furthermore, inference and learning can proceed in a biologically plausible fashion. Cortical responses can be seen as the brain’s attempt to minimize the free energy induced by a stimulus and thereby encode the most likely cause of that stimulus. Similarly, learning emerges from changes in synaptic efficacy that minimize the free energy, averaged over all stimuli encountered. The underlying scheme rests on empirical Bayes and hierarchical models of how sensory input is caused. The use of hierarchical models enables the brain to construct prior expectations in a dynamic and context-sensitive fashion. This scheme provides a principled way to understand many aspects of cortical organization and responses. The aim of this article is to encompass many apparently unrelated anatomical, physiological and psychophysical attributes of the brain within a single theoretical perspective.

In terms of cortical architectures, the theoretical treatment predicts that sensory cortex should be arranged hierarchically, that connections should be reciprocal and that forward and backward connections should show a functional asymmetry (forward connections are driving, whereas backward connections are both driving and modulatory). In terms of synaptic physiology, it predicts associative plasticity and, for dynamic models, spike-timing-dependent plasticity. In terms of electrophysiology, it accounts for classical and extra classical receptive field effects and long-latency or endogenous components of evoked cortical responses. It predicts the attenuation of responses encoding prediction error with perceptual learning and explains many phenomena such as repetition suppression, mismatch negativity (MMN) and the P300 in electroencephalography. In psychophysical terms, it accounts for the behavioural correlates of these physiological phenomena, for example, priming and global precedence. The final focus of this article is on perceptual learning as measured with the MMN and the implications for empirical studies of coupling among cortical areas using evoked sensory responses.

References

Absher J.R& Benson D.F. 1993Disconnection syndromes: an overview of Geschwind's contributions. Neurology. 43, 862–867. Crossref, PubMed, ISI, Google Scholar
Angelucci A, Levitt J.B& Lund J.SAnatomical origins of the classical receptive field and modulatory surround field of single neurons in macaque visual cortical area V1. Prog. Brain Res. 136, 2002a373–388. Crossref, PubMed, ISI, Google Scholar
Angelucci A, Levitt J.B, Walton E.J, Hupe J.M, Bullier J& Lund J.SCircuits for local and global signal integration in primary visual cortex. J. Neurosci. 22, 2002b8633–8646. Crossref, PubMed, ISI, Google Scholar
Atick J.J& Redlich A.N. 1990Towards a theory of early visual processing. Neural Comput. 2, 308–320. Crossref, ISI, Google Scholar
Baldeweg T, Klugman A, Gruzelier J.H& Hirsch S.R. 2002Impairment in frontal but not temporal components of mismatch negativity in schizophrenia. Int. J. Psychophysiol. 43, 111–122. Crossref, PubMed, ISI, Google Scholar
Baldeweg T, Klugman A, Gruzelier J& Hirsch S.R. 2004Mismatch negativity potentials and cognitive impairment in schizophrenia. Schizophr. Res. 6, 203–217. Crossref, ISI, Google Scholar
Ballard D.H, Hinton G.E& Sejnowski T.J. 1983Parallel visual computation. Nature. 306, 21–26. Crossref, PubMed, ISI, Google Scholar
Barlow H.BPossible principles underlying the transformation of sensory messages. Sensory communication& Rosenblith W.A. 1961Cambridge, MA:MIT Press. Google Scholar
Bell A.J& Sejnowski T.J. 1995An information maximisation approach to blind separation and blind de-convolution. Neural Comput. 7, 1129–1159. Crossref, PubMed, ISI, Google Scholar
Brodmann K. 1905Beitrage zur histologischen lokalisation der Groβhirnrinde. III. Mitteilung. Die Rindefelder der niederen Affen. J. Psychol. Neurol. 4, 177–226. Google Scholar
Brodmann KVergleichende Lokaisationslehre der Groβhirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues. 1909Leipzig:Barthpp. 1–9. Google Scholar
Büchel C& Friston K.J. 1997Modulation of connectivity in visual pathways by attention: cortical interactions evaluated with structural equation modelling and fMRI. Cereb. Cortex. 7, 768–778. Crossref, PubMed, ISI, Google Scholar
Buonomano D.V& Merzenich M.M. 1998Cortical plasticity: from synapses to maps. Annu. Rev. Neurosci. 21, 149–186. Crossref, PubMed, ISI, Google Scholar
Crick F& Koch C. 1998Constraints on cortical and thalamic projections: the no-strong-loops hypothesis. Nature. 391, 245–250. Crossref, PubMed, ISI, Google Scholar
David O& Friston K.J. 2003A neural mass model for MEG/EEG: coupling and neuronal dynamics. NeuroImage. 20, 1743–1755. Crossref, PubMed, ISI, Google Scholar
David O, Kiebel S.J, Harrison L.M, Mattout J, Kilner J.M& Friston K.JDynamic causal modelling of evoked responses in EEG and MEG. (2005). Google Scholar
Dayan P& Abbot L.FTheoretical neuroscience. Computational and mathematical modelling of neural systems. 2001Cambridge, MA:MIT Press. Google Scholar
Dayan P, Hinton G.E& Neal R.M. 1995The Helmholtz machine. Neural Comput. 7, 889–904. Crossref, PubMed, ISI, Google Scholar
Dempster A.P, Laird N.M& Rubin . 1977Maximum likelihood from incomplete data via the EM algorithm. J. R. Stat. Soc. Ser. B. 39, 1–38. Google Scholar
Desimone R. 1996Neural mechanisms for visual memory and their role in attention. Proc. Natl Acad. Sci. USA. 93, 13 494–13 499. Crossref, ISI, Google Scholar
Efron B& Morris C. 1973Stein's estimation rule and its competitors—an empirical Bayes approach. J. Am. Stat. Assoc. 68, 117–130. ISI, Google Scholar
Felleman D.J& Van Essen D.C. 1991Distributed hierarchical processing in the primate cerebral cortex. Cereb. Cortex. 1, 1–47. Crossref, PubMed, ISI, Google Scholar
Foldiak P. 1990Forming sparse representations by local anti-Hebbian learning. Biol. Cybern. 64, 165–170. Crossref, PubMed, ISI, Google Scholar
Frey U& Morris R.G.M. 1997Synaptic tagging and long-term potentiation. Nature. 385, 533–536. Crossref, PubMed, ISI, Google Scholar
Friston K.J. 1998The disconnection hypothesis. Schizophr. Res. 30, 115–125. Crossref, PubMed, ISI, Google Scholar
Friston K.J. 2000The labile brain. III. Transients and spatio-temporal receptive fields. Phil. Trans. R. Soc. B. 355, 253–265. Link, ISI, Google Scholar
Friston K.J. 2002Functional integration and inference in the brain. Prog. Neurobiol. 68, 113–143. Crossref, PubMed, ISI, Google Scholar
Friston K.J. 2003Learning and inference in the brain. Neural Netw. 16, 1325–1352. Crossref, PubMed, ISI, Google Scholar
Friston K.J, Harrison L& Penny W. 2003Dynamic causal modelling. NeuroImage. 19, 1273–1302. Crossref, PubMed, ISI, Google Scholar
Fuhrmann G, Segev I, Markram H& Tsodyks M. 2002Coding of temporal information by activity-dependent synapses. J. Neurophysiol. 87, 140–148. Crossref, PubMed, ISI, Google Scholar
Girard P& Bullier J. 1989Visual activity in area V2 during reversible inactivation of area 17 in the macaque monkey. J. Neurophysiol. 62, 1287–1301. Crossref, PubMed, ISI, Google Scholar
Han S& He X. 2003Modulation of neural activities by enhanced local selection in the processing of compound stimuli. Hum. Brain Mapp. 19, 273–281. Crossref, PubMed, ISI, Google Scholar
Harrison L.M, Rees G& Friston K.JExtra-classical and predictive coding effects measured with fMRI. (2004). Google Scholar
Harth E, Unnikrishnan K.P& Pandya A.S. 1987The inversion of sensory processing by feedback pathways: a model of visual cognitive functions. Science. 237, 184–187. Crossref, PubMed, ISI, Google Scholar
Helmholtz H& Southall J.P.C vol. 31860/1962New York:Dover(English trans.). Google Scholar
Henson R, Shallice T& Dolan R. 2000Neuroimaging evidence for dissociable forms of repetition priming. Science. 287, 1269–1272. Crossref, PubMed, ISI, Google Scholar
Hilgetag C.C, O'Neill M.A& Young M.P. 2000Hierarchical organisation of macaque and cat cortical sensory systems explored with a novel network processor. Phil. Trans. R. Soc. B. 355, 71–89. Link, ISI, Google Scholar
Hinton G.E, Dayan P, Frey B.J& Neal R.M. 1995The ‘Wake–Sleep’ algorithm for unsupervised neural networks. Science. 268, 1158–1161. Crossref, PubMed, ISI, Google Scholar
Hirsch J.A& Gilbert C.D. 1991Synaptic physiology of horizontal connections in the cat's visual cortex. J. Neurosci. 11, 1800–1809. Crossref, PubMed, ISI, Google Scholar
Hochstein S& Ahissar M. 2002View from the top: hierarchies and reverse hierarchies in the visual system. Neuron. 36, 791–804. Crossref, PubMed, ISI, Google Scholar
Hupe J.M, James A.C, Payne B.R, Lomber S.G, Girard P& Bullier J. 1998Cortical feedback improves discrimination between figure and background by V1, V2 and V3 neurons. Nature. 394, 784–787. Crossref, PubMed, ISI, Google Scholar
Jääskeläinen I.P, et al.2004Human posterior auditory cortex gates novel sounds to consciousness. Proc. Natl Acad. Sci. 101, 6809–6814. Crossref, PubMed, ISI, Google Scholar
Jansen B.H& Rit V.G. 1995Electroencephalogram and visual evoked potential generation in a mathematical model of coupled cortical columns. Biol. Cybern. 73, 357–366. Crossref, PubMed, ISI, Google Scholar
Kass R.E& Steffey D. 1989Approximate Bayesian inference in conditionally independent hierarchical models (parametric empirical Bayes models). J. Am. Stat. Assoc. 407, 717–726. Crossref, Google Scholar
Kawato M, Hayakawa H& Inui T. 1993A forward-inverse optics model of reciprocal connections between visual cortical areas. Network. 4, 415–422. Crossref, ISI, Google Scholar
Kay J& Phillips W.A. 1996Activation functions, computational goals and learning rules for local processors with contextual guidance. Neural Comput. 9, 895–910. Crossref, ISI, Google Scholar
Kepecs A, Van Rossum M.C, Song S& Tegner J. 2002Spike-timing-dependent plasticity: common themes and divergent vistas. Biol. Cybern. 87, 446–458. Crossref, PubMed, ISI, Google Scholar
Kersten D, Mamassian P& Yuille A. 2004Object perception as Bayesian inference. Annu. Rev. Psychol. 55, 271–304. Crossref, PubMed, ISI, Google Scholar
Kötter R& Wanke E. 2005Mapping brains without coordinates. Phil. Trans. R. Soc. B. 360, . Link, ISI, Google Scholar
Lee T.S& Mumford D. 2003Hierarchical Bayesian inference in the visual cortex. J. Opt. Soc. Am. Opt. Image Sci. Vis. 20, 1434–1448. Crossref, PubMed, ISI, Google Scholar
Linsker R. 1990Perceptual neural organisation: some approaches based on network models and information theory. Annu. Rev. Neurosci. 13, 257–281. Crossref, PubMed, ISI, Google Scholar
Locke JAn essay concerning human understanding. 1690/1976London:Dent. Google Scholar
MacKay D.MThe epistemological problem for automata. In. Automata studies (ed. C.E. Shannon & J. McCarthy)1956Princeton, NJ:Princeton University Presspp. 235–251. Google Scholar
Martin S.J, Grimwood P.D& Morris R.G. 2000Synaptic plasticity and memory: an evaluation of the hypothesis. Annu. Rev. Neurosci. 23, 649–711. Crossref, PubMed, ISI, Google Scholar
Maunsell J.H& Van Essen D.C. 1983The connections of the middle temporal visual area (MT) and their relationship to a cortical hierarchy in the macaque monkey. J. Neurosci. 3, 2563–2586. Crossref, PubMed, ISI, Google Scholar
Mehta M.R. 2001Neuronal dynamics of predictive coding. Neuroscientist. 7, 490–495. Crossref, PubMed, ISI, Google Scholar
Mesulam M.M. 1998From sensation to cognition. Brain. 121, 1013–1052. Crossref, PubMed, ISI, Google Scholar
Mumford D. 1992On the computational architecture of the neocortex. II. The role of cortico-cortical loops. Biol. Cybern. 66, 241–251. Crossref, PubMed, ISI, Google Scholar
Murphy P.C& Sillito A.M. 1987Corticofugal feedback influences the generation of length tuning in the visual pathway. Nature. 329, 727–729. Crossref, PubMed, ISI, Google Scholar
Murray S.O, Kersten D, Olshausen B.A, Schrater P& Woods D.L. 2002Shape perception reduces activity in human primary visual cortex. Proc. Natl Acad. Sci. USA. 99, 15 164–15 169. Crossref, ISI, Google Scholar
Näätänen R. 2003Mismatch negativity: clinical research and possible applications. Int. J. Psychophysiol. 48, 179–188. Crossref, PubMed, ISI, Google Scholar
Näätänen R, Pakarinen S, Rinne T& Takegata R. 2004The mismatch negativity (MMN): towards the optimal paradigm. Clin. Neurophysiol. 115, 140–144. Crossref, PubMed, ISI, Google Scholar
Neisser UCognitive psychology. 1967New York:Appleton-Century-Crofts. Google Scholar
Oja E. 1989Neural networks, principal components, and subspaces. Int. J. Neural Syst. 1, 61–68. Crossref, Google Scholar
Olshausen B.A& Field D.J. 1996Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature. 381, 607–609. Crossref, PubMed, ISI, Google Scholar
Optican L& Richmond B.J. 1987Temporal encoding of two-dimensional patterns by single units in primate inferior cortex. II. Information theoretic analysis. J. Neurophysiol. 57, 132–146. Crossref, PubMed, ISI, Google Scholar
Pack C.C& Born R.T. 2001Temporal dynamics of a neural solution to the aperture problem in visual area MT of macaque brain. Nature. 409, 1040–1042. Crossref, PubMed, ISI, Google Scholar
Phillips W.A& Singer W. 1997In search of common foundations for cortical computation. Behav. Brain Sci. 20, 57–83. Crossref, PubMed, ISI, Google Scholar
Phillips C.G, Zeki S& Barlow H.B. 1984Localization of function in the cerebral cortex; past present and future. Brain. 107, 327–361. Crossref, PubMed, ISI, Google Scholar
Poggio T, Torre V& Koch C. 1985Computational vision and regularization theory. Nature. 317, 314–319. Crossref, PubMed, ISI, Google Scholar
Pollen D.A. 1999On the neural correlates of visual perception. Cereb. Cortex. 9, 4–19. Crossref, PubMed, ISI, Google Scholar
Rainer G, Rao S.C& Miller E.K. 1999Prospective coding for objects in primate prefrontal cortex. J. Neurosci. 19, 5493–5505. Crossref, PubMed, ISI, Google Scholar
Rao R.P. 1999An optimal estimation approach to visual perception and learning. Vision Res. 39, 1963–1989. Crossref, PubMed, ISI, Google Scholar
Rao R.P& Ballard D.H. 1999Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive field effects. Nat. Neurosci. 2, 79–87. Crossref, PubMed, ISI, Google Scholar
Rivadulla C, Martinez L.M, Varela C& Cudeiro J. 2002Completing the corticofugal loop: a visual role for the corticogeniculate type 1 metabotropic glutamate receptor. J. Neurosci. 22, 2956–2962. Crossref, PubMed, ISI, Google Scholar
Rockland K.S& Pandya D.N. 1979Laminar origins and terminations of cortical connections of the occipital lobe in the rhesus monkey. Brain Res. 179, 3–20. Crossref, PubMed, ISI, Google Scholar
Salin P.-A& Bullier J. 1995Corticocortical connections in the visual system: structure and function. Psychol. Bull. 75, 107–154. Google Scholar
Sandell J.H& Schiller P.H. 1982Effect of cooling area 18 on striate cortex cells in the squirrel monkey. J. Neurophysiol. 48, 38–48. Crossref, PubMed, ISI, Google Scholar
Sherman S.M& Guillery R.W. 1998On the actions that one nerve cell can have on another: distinguishing ‘drivers’ from ‘modulators’. Proc. Natl Acad. Sci. USA. 95, 7121–7126. Crossref, PubMed, ISI, Google Scholar
Simoncelli E.P& Olshausen B.A. 2001Natural image statistics and neural representation. Annu. Rev. Neurosci. 24, 1193–1216. Crossref, PubMed, ISI, Google Scholar
Sugase Y, Yamane S, Ueno S& Kawano K. 1999Global and fine information coded by single neurons in the temporal visual cortex. Nature. 400, 869–873. Crossref, PubMed, ISI, Google Scholar
Tononi G, Sporns O& Edelman G.M. 1994A measure for brain complexity: relating functional segregation and integration in the nervous system. Proc. Natl Acad. Sci. USA. 91, 5033–5037. Crossref, PubMed, ISI, Google Scholar
Tovee M.J, Rolls E.T, Treves A& Bellis R.P. 1993Information encoding and the response of single neurons in the primate temporal visual cortex. J. Neurophysiol. 70, 640–654. Crossref, PubMed, ISI, Google Scholar
Umbricht D, Schmid L, Koller R, Vollenweider F.X, Hell D& Javitt D.C. 2000Ketamine-induced deficits in auditory and visual context-dependent processing in healthy volunteers. Arch. Gen. Psychiatry. 57, 1139–1147. Crossref, PubMed, Google Scholar
Zeki SThe motion pathways of the visual cortex. Vision: coding and efficiency& Blakemore C. 1990pp. 321–345. Eds. UK:Cambridge University Press. Google Scholar
Zeki SA vision of the brain. 1993Oxford:Blackwell Scientific. Google Scholar
Zeki S& Shipp S. 1988The functional logic of cortical connections. Nature. 335, 311–317. Crossref, PubMed, ISI, Google Scholar