Neuronal tuning

From Wikipedia, the free encyclopedia

Neuronal tuning refers to the property of brain cells to selectively represent a particular kind of sensory, motor, or cognitive information. For example, an auditory system neuron best responding to the sound of particular frequency is said to be tuned to that frequency. In the visual system, there are neurons tuned to particular objects, for example faces. In the motor system, tuning to movement direction has been described. Neuronal tuning is not necessarily sharp. Rather, it can be quite broad. In the latter case, neuroscientists often argue that information is represented in distributed manner by neural ensembles. In an ensemble, individual elements can be weakly tuned, but the ensemble as a whole still encodes the information.

One neuron can be tuned to several parameters. Examples include neurons in association areas of cerebral cortex that respond to stimuli of different modality (for example, visual stimulation, touch, sound). The field of brain-computer interface makes use of such neurons by reading-out different kind of information from their firing, for example simultaneous decoding of hand velocity and hand gripping force. In prefrontal cortex, neurons have been described that simultaneously encode information about the orientation of selective visual attention and short-term visual memory (Attention versus memory in prefrontal cortex). Neurons that are tuned to different signals are often said to integrate information from different sources. In neural networks, such integration is the major principle of operation.