Neural selectivity in the Cocktail Party Problem

Details

Published on 12 April 2013

The inability to have a conversation in a crowded room is one of the chief complaints among those with hearing loss. Now for the first time, researchers have demonstrated how the brain can hone in on a single voice in a sea of speakers.

“There’s no way to ‘close your ear’ so that all the sounds in the environment are represented in the brain, at least at the sensory level,” says Columbia University psychiatrist Charles Schroeder, who was the lead author on the study.

Using the “Cocktail Party” paradigm, along with direct recordings from the cortical surface in surgical epilepsy patients, the researchers discovered that brain activity dynamically tracks speech streams using both low-frequency phase and high-frequency amplitude fluctuations, and that the optimal encoding likely combines the two.

In and near low-level auditory cortices, they found that attention “modulates” the representation by enhancing cortical tracking of attended speech streams, but ignored speech remains represented. In higher-order regions, the representation appears to be more “selective,” in that there is no detectable tracking of ignored speech. This selectivity itself, they says, seems to sharpen as a sentence unfolds.

The authors found that both low-frequency phase and high-gamma power preferentially track attended speech; near auditory cortex attention modulates response to attended and ignored speakers; selective tracking of only the attended talker increases in higher-order regions; and selectivity for the attended speaker increases over time.

Source: Neuron, Volume 77, Issue 5, 6 March 2013