Encoding information into autonomously bursting neural network with pairs of time-delayed pulses

June Hoan Kim, Ho Jun Lee, Wonshik Choi, Kyoung Jin Lee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Biological neural networks with many plastic synaptic connections can store external input information in the map of synaptic weights as a form of unsupervised learning. However, the same neural network often produces dramatic reverberating events in which many neurons fire almost simultaneously – a phenomenon coined as ‘population burst.’ The autonomous bursting activity is a consequence of the delicate balance between recurrent excitation and self-inhibition; as such, any periodic sequences of burst-generating stimuli delivered even at a low frequency (~1 Hz) can easily suppress the entire network connectivity. Here we demonstrate that ‘Δt paired-pulse stimulation’, can be a novel way for encoding spatially-distributed high-frequency (~10 Hz) information into such a system without causing a complete suppression. The encoded memory can be probed simply by delivering multiple probing pulses and then estimating the precision of the arrival times of the subsequent evoked recurrent bursts.

Original languageEnglish
Article number1394
JournalScientific Reports
Volume9
Issue number1
DOIs
Publication statusPublished - 2019 Dec 1

    Fingerprint

ASJC Scopus subject areas

  • General

Cite this