The Neurons That Translate Noises Into Brain Code


One of the brain’s fundamental transformations of information has been observed by University of Oregon scientists. And what they found in the brains of rats, they report, may be true across all mammals.

When you hear footsteps or a woodpecker’s drilling, the rhythmic structure of the sounds is striking, says Michael Wehr, a professor of psychology at the University of Oregon. Even when the temporal structure of a sound is less obvious, as with human speech, the timing still conveys important information:

“As the information progresses towards the auditory cortex, however, the representation of sound undergoes a transformation. There is a gradual shift towards neurons that use an entirely different system for encoding information.”

The findings reflect those previously shown in primates.

Brain Languages

Neurons in the brain use two different languages to encode information: temporal coding and rate coding.

For neurons in the auditory thalamus, the part of the brain that relays information from the ears to the auditory cortex, this takes the form of temporal coding.

Neurons fire in sync with the original sound, providing an exact replication of the sound’s structure in time.

In the auditory cortex, about half the neurons use rate coding. The structure of the sound is conveyed through the density and rate of the neurons’ spiking, rather than the exact timing.

To study how the coding system switch, Wehr and undergraduate student Xiang used whole-cell recording to capture thousands of interactions within a single neuron each time it responds to a sound.

The team saw how individual cells respond to a stream of rhythmic clicks, and noted that individual rate-coding neurons received up to 82 percent of their inputs from temporal-coding neurons.

“This means that these neurons are acting like translators, converting a sound from one language to another,” Wehr says. “By peering inside a neuron, we can see the mechanism for how the translation is taking place.”

Deep Brain Circuits

One mechanism is the way that excitatory and inhibitory neurons cooperate to push and pull together, like parents pushing a playground teeter-totter, he says.

In response to a click, excitatory neurons push on a cell and inhibitory neurons follow with a pull exactly out of phase with the excitatory neurons. The combination drives cells to fire spikes at a high rate, converting the temporal code into a rate code.

The observation provides a glimpse into how circuits deep in the brain give rise to how the world is perceived, Wehr says.

Neuroscientists have speculated that the transformation from temporal coding to rate coding may explain the perceptual boundary experienced between rhythm and pitch. Slow trains of clicks sound rhythmic, but fast trains of clicks sound like a buzzy tone.

It could be that the different experiences of sound are produced by the two different kinds of neurons, Wehr says. It could be, he adds, that rate coding is a universal language that helps put together what people see and hear.


Xiang Gao, Michael Wehr
A Coding Transformation for Temporally Structured Sounds within Auditory Cortical Neurons
Neuron, Volume 86, Issue 1, 8 April 2015, Pages 292–303 doi:10.1016/j.neuron.2015.03.004