Submitted on November 21, 2008
Revised on March 18, 2009
Accepted on March 18, 2009
Wiener-Volterra Characterization of Neurons in Primary Auditory Cortex Using Poisson-Distributed Impulse Train Inputs
Martin Pienkowski1, Gregory Shaw2, and Jos J Eggermont2*
1 Univerity of Calgary
2 University of Calgary
* To whom correspondence should be addressed. E-mail: eggermon{at}ucalgary.ca.
An extension of the Wiener-Volterra theory to a Poisson-distributed impulse train input was used to characterize neurons in primary auditory cortex (AI) of the ketamine-anesthetized cat. Both linear and second-order "Poisson-Wiener" (PW) models were tested on their predictions of AI temporal modulation transfer functions (tMTFs) obtained with periodic click trains. Second-order PW models almost always proved superior to linear models, and typically gave good fits to measured tMTFs. Second-order kernels invariably demonstrated compressive nonlinearities in the AI neuron response. In neurons with lowpass tMTFs, the strength of the compression decayed exponentially with the interstimulus lag, whereas in neurons with bandpass tMTFs, a distinct secondary peak in compression strength was usually seen at a lag that correlated with the neuron's best modulation frequency. It is suggested that modulation tuning in AI arises in part from an interplay of two nonlinear synaptic processes with distinct time courses.
