Publications

We characterize primary auditory cortex (AI) units using a neural model for the detection of frequency and amplitude transitions. The model is a generalization of a model for the detection of amplitude transition. A set of neurons, tuned in the spectrotemporal domain, is created by means of neural delays and frequency filtering. The sensitivity of the model to frequency and amplitude transitions is achieved by applying a 2-dimensional rotatable receptive field to the set of spectrotemporally tuned neurons. We evaluated the model using data recorded in AI of anesthetized ferrets. We show that the model is able to fit the responses of AI units to variety of stimuli, including single tones, delayed 2-tone stimuli and various frequency-modulated tones, using only a small number of parameters. Furthermore, we show that the topographical order in maps of the model parameters is higher than in maps created from response indices extracted directly from the responses to any single stimulus. These results suggest a possible ordered organization of a simple rotatable spectrotemporal receptive field in the mammalian AI.