Neural mechanisms of stream formation during active listening in the ferret auditory cortex

Listening in the real world involves making sense of mixtures of multiple overlapping sounds. The brain decomposes such scenes into individual objects, and a sequence of related auditory objects forms a stream. We are investigating the role of the auditory cortex in the formation and maintenance of auditory streams. The temporal coherence theory has provided one explanation for stream formation, postulating that the brain creates a multidimensional representation of sounds along different feature axes, and groups them based on their temporal coherence, to form streams. However, this theory has yet to be tested at the neural population level with naturalistic sounds. To investigate this question, we trained ferrets to detect a target word in a stream of random distractor words, spoken by the same talker, played in the presence of a spatially separated noise stream (speech-shaped noise). Neural data were collected in the auditory cortex of behaving ferrets. We found that the neural encoding of words is unaffected by the location of the stream in a single-stream task (p > 0.05). However, in the presence of a spatially separated competing stream, encoding of the words in the contralateral stream was significantly better than in the ipsilateral stream (p = 0.034). Interestingly, early in the trial, the ipsilateral stream decoding is at chance level, whereas towards the end of the trial it approaches that of the contralateral stream. These findings suggest that, in the presence of a competing noise stream, auditory cortical neural populations may quickly adapt to switch from representing the contralateral stream to representing the attended stream.