Abstract
We modeled suppressive effect on visual perception induced by TMS. TMS is widely used in experimental studies about visual perception, however neural mechanisms underlying TMS interference are still unclear, especially in theoretical perspective. Here we used the simplest excitatory-inhibitory balanced network showing orientation selectivity in V1 as a model of neural population and analyzed the response to a TMS-like perturbation, simulating the fundamental property that TMS briefly and simultaneously stimulates neural population in local cortical area under the coil. Applying the perturbation, mean activity of the network transiently increased and then decreased for a longer period followed by a loss of an orientation tuning profile. If afferent input had a large transient and weak sustained component, there was a critical latency period during which the perturbation could completely suppress the network activity. The range of the suppressive latency period increased with decrease of afferent intensity and reached over 100ms if the afferent intensity approached excitation threshold of the network. These results well agree with typical experimental data of visual suppression by occipital TMS. In occipital TMS experiments, applying multiple pulses can facilitate phosphene or suppression even if each single pulse cannot induce any perceptible effects. Such subthreshold accumulation was also observed in the network model in comparable time range to experimental data, but not in isolated single neuron model. In the network model, a subthreshold conditioning stimulus could induce sustained inhibitory bias so that the following suppressive threshold was decreased and the range of suppressive latency period was prolonged, which are also parallel to experimental data of occipital paired TMS. These results suggest that, in addition to effect on a single neuron, inhibitory interaction in neural population plays an important role in TMS-induced visual suppression.