Integration over multiple timescales in primary auditory cortex

Stephen V. David, Shihab A. Shamma

Research output: Contribution to journalArticle

34 Scopus citations

Abstract

Speech and other natural vocalizations are characterized by large modulations in their sound envelope. The timing of these modulations contains critical information for discrimination of important features, such as phonemes.Westudied how depression of synaptic inputs, a mechanism frequently reported in cortex, can contribute to the encoding of envelope dynamics. Using a nonlinear stimulus-response model that accounted for synaptic depression, we predicted responses of neurons in ferret primary auditory cortex (A1) to stimuli with natural temporal modulations. The depression model consistently performed better than linear and second-order models previously used to characterize A1 neurons, and it produced more biologically plausible fits. To test how synaptic depression can contribute to temporal stimulus integration, we used nonparametric maximum a posteriori decoding to compare the ability of neurons showing and not showing depression to reconstruct the stimulus envelope. Neurons showing evidence for depression reconstructed stimuli over a longer range of latencies. These findings suggest that variation in depression across the cortical population supports a rich code for representing the temporal dynamics of natural sounds.

Original languageEnglish (US)
Pages (from-to)19154-19166
Number of pages13
JournalJournal of Neuroscience
Volume33
Issue number49
DOIs
StatePublished - Dec 9 2013

    Fingerprint

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this