Postontogenetic short-term plasticity in the somatosensory system: a neural network model

Both repetitive tactile stimulation applied to the hindpaw of an adult rat and intracortical microstimulation in the primary somatosensory cortex lead to fully reversible plastic changes in the cortical hindpaw representation, which occur within hours. Psychophysical experiments in humans indicate that the plastic changes are accompanied by an increase in tactile resolution. We introduce a biologically oriented neural network to simulate the different experiments. The model uses a local learning rule which accounts for all major effects observed, such as the dynamical changes of receptive field size, the increase of the cortical representation area and the full reversibility of all plastic changes. We also discuss the conditions under which an increase in receptive field size of somatosensory neurons results in a higher sensory resolution