288 pp. per issue
6 x 9, illustrated
2014 Impact factor:

Neural Computation

May 2017, Vol. 29, No. 5, Pages 1263-1292
(doi: 10.1162/NECO_a_00952)
© 2017 Massachusetts Institute of Technology
Modulation of Context-Dependent Spatiotemporal Patterns within Packets of Spiking Activity
Article PDF (750.32 KB)

Recent experiments have shown that stereotypical spatiotemporal patterns occur during brief packets of spiking activity in the cortex, and it has been suggested that top-down inputs can modulate these patterns according to the context. We propose a simple model that may explain important features of these experimental observations and is analytically tractable. The key mechanism underlying this model is that context-dependent top-down inputs can modulate the effective connection strengths between neurons because of short-term synaptic depression. As a result, the degree of synchrony and, in turn, the spatiotemporal patterns of spiking activity that occur during packets are modulated by the top-down inputs. This is shown using an analytical framework, based on avalanche dynamics, that allows calculating the probability that a given neuron spikes during a packet and numerical simulations. Finally, we show that the spatiotemporal patterns that replay previously experienced sequential stimuli and their binding with their corresponding context can be learned because of spike-timing-dependent plasticity.