Neural encoding of natural sounds in the auditory system

Information theoretic related measures are developed to investigate the encoding of natural sounds in the auditory system of the songbird. Neural response variability can be characterized by the coherence and correlation coefficient between neural responses and their time-varying mean response. We introduce novel un-biased estimators for these measures and show how they can be used to validate stimulus response models in such a way that inaccuracy due to model limitations can be separated from inaccuracy due to the variability in the neural response. Neural responses also can be characterized by their ability to discriminate among different classes of stimuli. Songbirds are highly capable of processing complex natural sounds (birdsong). We investigate the hypothesis that the auditory neurons of songbirds are tuned for the processing of birdsong. The ability of sensory auditory neurons to convey information about the world can be quantified by the mutual information between neural responses and stimuli. We analyzed neural responses to natural birdsong and two synthetic stimuli: synthetic songs that matched the power of spectral-temporal modulations but lacked the modulation phase structure of zebra finch song and noise with uniform band-limited spectral-temporal modulations. Using a novel parametric method for estimating mutual information with limited data, we compared responses to these natural and synthetic stimuli. By defining neural selectivity as relative mutual information, we found that the auditory system of songbirds showed selectivity for song and song-like sounds. We also extend our analysis of single neuron responses to the discrimination capabilities in the ensemble neural response. We quantified our ensemble response properties with an extension of a simple distance measure, which we also used to quantify the neural discrimination of single neurons. The results of our analysis showed greatest ensemble neural discrimination for natural stimuli.