Lateral Inhibition

The nervous system uses lateral inhibition to improve spatial resolution and contrast. Suppose, for example, that we have a stimulus distribution shaped like a bell. In lateral inhibition the stimulus distribution is shifted laterally (left and right in this case, say) by lateral branches of afferent axons, and subtracted (hence the designation inhibition) from the original stimulus curve. This yields a narrower curve (the sides of the ?>bell?> are steeper).

Suppose that we have two bell-shaped stimulus curves, so close together that they partially merge to yield a single stimulus peak. Lateral inhibition may nevertheless be able to reveal that two stimuli are actually present.

A hypothetical three-stage model is examined in which the primary stimulus is a blunt ?>compass?> point pressing against the hand, and lateral inhibition is applied in the spinal cord, thalamus, and somatosensory cortex.

Several two-dimensional models are also examined. A special case known as zero-sum lateral inhibition is especially important because it can extract the edges of the spatial stimulus curve.