HomeNewIntroductionQuantum Mind BlogQuantum Mind TheoriesRelated TopicsKey ArticlesReferencesContact UsOnline Book

Gamma synchrony and consciousness




Neocortical Rhythms

Wolf Singer

In:- Dynamic Coordination in the Brain Eds. Christoph von der Malsburg, William Phillips, & Wolf Singer

INTRODUCTION:  Singer's chapter looks to come near to finalising the case for a correlation between the gamma synchrony and consciousness. Conscious stimuli are associated with phase locking of gamma oscillations across spatially distributed regions of the cortex, and also with increases in synchrony without increases in rate of firing. In terms of perception, neurons responding to eyes, noses etc can synchronise to identify a face, but when the facial components are scrambled the synchrony, but not the discharge strength, disappears.

This chapter discusses the rhythmic modulation of neuronal activity. During processing in the cortex, the brain increasingly selects for the relationship between objects. This involves interactions between different parts of the cortex. There is a requirement to cope with the ambiguity of the external world. The environment may contain objects with contours that overlap, or are partly hidden, and these conflicting signals have to be resolved in the cortex. Further to this some objects are encoded in different sensory modalities. Evidence suggests that this process involves not only individual neurons but also assemblies of neurons (1. Singer, 1999, 2. Tsunoda et al, 2001). The possible conjunctions in perception are too large to be dealt with by individual neurons, and utilise assemblies of neurons with each neuron relating to particular aspects of the object.

There appear to be two stages to this process. There is a signal to indicate that certain features are present. This operates on a 'rate code' basis, where a higher discharge frequency codes for a greater probability of a particular feature being present. However, evidence (3. Gray et al, 1989) shows that neurons in the primary visual cortex synchronise their spiking, and it is proposed that synchronisation of responses could code from relatedness. Synchronised inputs to neurons have a stronger impact than unsynchronised signals. This applies not only to the processing of perception, but also to learning, where precise temporal relations between pre and post-synaptic activity leads to strengthening of synapses. Signal processing and learning thus appear to be related processes.

Synchronisation of discharges is often associated with oscillations in populations of neurons. These oscillations are driven by inhibitory neurons through both synapses and gap junctions. (4. Kopell et al, 2000, 5. Whittington et al, 2001) Inhibitory inputs to pyramidal cells favour discharges at depolarising peaks and this allows synchrony in firing. Locally synchronised oscillations can become phase locked with others that are spatially separated with zero delay. Oscillations in different frequencies can exist together, and this combination may deal with composite objects or movement trajectories. Synchronisations may be used at all stages of neural processing from the retina through to the cortex. This appears to serve the function of creating a relationship between spatially distributed responses, for instance signalling a relationship between neuron groupings that may be utilised in future processing, notably perceptual grouping.

Synchronisation involves oscillations at up to 90 Hz in both the beta and gamma range. This synchronisation is related to connections linking cortical columns encoding for linked features. The inferior temporal cortex is regarded as the likely site for the production of visual objects, and object-related assemblies are associated with synchronisation. In one study, neurons responding to eyes, noses and faces were shown to synchronise to recognise a face. If the individual components were scrambled into a non-face arrangement then synchrony did not arise. However, the scrambling into non-face did not alter the discharge rate, only the synchrony. Focus of attention on objects also caused increased synchrony in the beta and gamma bands. Here again synchronisation does not necessarily relate to increased discharge rates. Coherent oscillations across distributed areas of the cortex including executive areas are seen as facilitating the routing of activity and the rapid response of different areas of the cortex.

Evidence also indicates a close correlation between gamma synchronisation and conscious processing (6. Melloni et al, 2007). Activity related to conscious responses is more synchronised, but not more vigorous. In human subjects conscious processing has been related to phase locked gamma oscillations in widely distributed cortical areas, whereas unconscious processing produced only local gamma activity (6. Melloni et al, 2007).

References:-
1.) Singer, W. (1999)  -  Neuronal synchrony: A versatile code for the definition of relations  -  Neuron, 24 (1), pp. 49-65
2.)  Tsunoda, K. et al (2001)  -  Complex objects represented in inferotemporal cortex by the recombination of feature columns  -  Nature Neuroscience, 4 (8): pp. 832-8
3.)  Gray et al (1989)  -  Oscillatory responses exhibit inter-columnar synchronisation which reflects global stimulus properties  -  Nature, 338: pp. 334-7
4.)  Koppel, N. et al (2000) - Gamma and beta rhythms have different synchronisation properties  -  PNAS, 97, (4): pp. 1867-72
5.) Whittington et al (2001)  -  Synaptic and non-synaptic mechanisms underlying stimulus induced gamma oscillations  -  Journal of Neuroscience, 21 (5): pp. 1727-1738 P. 6.)  Melloni, L. et al (2007)  -  Synchronisation of neural activity across cortical areas correlates with conscious perception  -  Journal of Neuroscience, 27 (11): pp. 2858-2865