Limits...
For motor outputs, as for sensory inputs, spike timing carries more information than spike rate.

Robinson R - PLoS Biol. (2014)

View Article: PubMed Central - PubMed

Affiliation: Freelance Science Writer, Sherborn, Massachusetts, United States of America.

AUTOMATICALLY GENERATED EXCERPT
Please rate it.

New work by Claire Tang, Samuel Sober, and colleagues in this issue of PLOS Biology indicates that, just as in sensory systems, the motor output that controls bird song is dictated not only by firing rate but also by firing pattern and the information inherent in those patterns greatly exceeds the information available in simple spike counts... Iterations of each song syllable were classified into “behavioral groups,” based on the acoustic properties of pitch, amplitude, and “spectral entropy,” which is the measure of noise versus tonality of a given sound... Electrodes in the RA allowed the authors to record spike trains that immediately preceded a syllable, thus enabling them to compare the firing parameters with the characteristics of the song it produced (Figure 1)... With the spike trains and acoustic analysis in hand, they next asked whether the firing rate (“spike counts”) or firing pattern (“spike timing”) of RA output was the best predictor of the behavioral group a given syllable would fall into... The final cost is the measure of dissimilarity, or “distance,” between the two trains... These distances are then used to independently classify the syllables they give rise to into behavioral groups, and those classifications are compared to the ones from the acoustic analysis... Next, the authors performed a second type of analysis, which examined the predictive information present in finer and finer timescale resolutions, from 40 milliseconds down to 1 millisecond, the shortest interval they could reliably measure... They found that as the temporal resolution increased, so did the information present, suggesting, in the authors' words, “that small differences in spike timing can significantly influence motor output. ” Indeed, the information present at that resolution was about ten times what was available from the spike rate alone... The results of these analyses have several implications... Most directly, restricting neural analysis to spike timing will likely significantly underestimate the information output of motor systems... Furthermore, while the spike trains analyzed here were from single neurons, the authors note that even higher-order information may be present in the coordinated output of multiple neurons... Finally, whether time-coded sensory inputs feed directly into time-coded motor outputs will require further investigation, but the demonstration of the existence of such outputs clears the way for those investigations. doi:10.1371/journal.pbio.1002018

Show MeSH
Millisecond-scale motor encoding in the songbird brain.Patterns of electrical activity in neurons (voltage traces, top) control skilled behaviors. Representing neural activity as binary “words” (middle) reveals how the brains of songbirds use these electrical patterns to produce vocal output. The precise timing of neural signals appears to be crucial in controlling the elaborate vocal patterns that have fascinated human listeners for centuries (bottom, excerpt from Olivier Messaien's Oiseaux exotiques, which was based on Messaien's transcriptions of birdsong). Image credit: Sam Sober.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4260779&req=5

pbio-1002019-g001: Millisecond-scale motor encoding in the songbird brain.Patterns of electrical activity in neurons (voltage traces, top) control skilled behaviors. Representing neural activity as binary “words” (middle) reveals how the brains of songbirds use these electrical patterns to produce vocal output. The precise timing of neural signals appears to be crucial in controlling the elaborate vocal patterns that have fascinated human listeners for centuries (bottom, excerpt from Olivier Messaien's Oiseaux exotiques, which was based on Messaien's transcriptions of birdsong). Image credit: Sam Sober.

Mentions: The basic unit of a bird's song is a note; one or more notes are combined to make a syllable, syllables are repeated to form a motif, and motifs together make a song. Iterations of each song syllable were classified into “behavioral groups,” based on the acoustic properties of pitch, amplitude, and “spectral entropy,” which is the measure of noise versus tonality of a given sound. Together, these three features capture most of the variation among the syllables produced by the finch. Electrodes in the RA allowed the authors to record spike trains that immediately preceded a syllable, thus enabling them to compare the firing parameters with the characteristics of the song it produced (Figure 1).


For motor outputs, as for sensory inputs, spike timing carries more information than spike rate.

Robinson R - PLoS Biol. (2014)

Millisecond-scale motor encoding in the songbird brain.Patterns of electrical activity in neurons (voltage traces, top) control skilled behaviors. Representing neural activity as binary “words” (middle) reveals how the brains of songbirds use these electrical patterns to produce vocal output. The precise timing of neural signals appears to be crucial in controlling the elaborate vocal patterns that have fascinated human listeners for centuries (bottom, excerpt from Olivier Messaien's Oiseaux exotiques, which was based on Messaien's transcriptions of birdsong). Image credit: Sam Sober.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4260779&req=5

pbio-1002019-g001: Millisecond-scale motor encoding in the songbird brain.Patterns of electrical activity in neurons (voltage traces, top) control skilled behaviors. Representing neural activity as binary “words” (middle) reveals how the brains of songbirds use these electrical patterns to produce vocal output. The precise timing of neural signals appears to be crucial in controlling the elaborate vocal patterns that have fascinated human listeners for centuries (bottom, excerpt from Olivier Messaien's Oiseaux exotiques, which was based on Messaien's transcriptions of birdsong). Image credit: Sam Sober.
Mentions: The basic unit of a bird's song is a note; one or more notes are combined to make a syllable, syllables are repeated to form a motif, and motifs together make a song. Iterations of each song syllable were classified into “behavioral groups,” based on the acoustic properties of pitch, amplitude, and “spectral entropy,” which is the measure of noise versus tonality of a given sound. Together, these three features capture most of the variation among the syllables produced by the finch. Electrodes in the RA allowed the authors to record spike trains that immediately preceded a syllable, thus enabling them to compare the firing parameters with the characteristics of the song it produced (Figure 1).

View Article: PubMed Central - PubMed

Affiliation: Freelance Science Writer, Sherborn, Massachusetts, United States of America.

AUTOMATICALLY GENERATED EXCERPT
Please rate it.

New work by Claire Tang, Samuel Sober, and colleagues in this issue of PLOS Biology indicates that, just as in sensory systems, the motor output that controls bird song is dictated not only by firing rate but also by firing pattern and the information inherent in those patterns greatly exceeds the information available in simple spike counts... Iterations of each song syllable were classified into “behavioral groups,” based on the acoustic properties of pitch, amplitude, and “spectral entropy,” which is the measure of noise versus tonality of a given sound... Electrodes in the RA allowed the authors to record spike trains that immediately preceded a syllable, thus enabling them to compare the firing parameters with the characteristics of the song it produced (Figure 1)... With the spike trains and acoustic analysis in hand, they next asked whether the firing rate (“spike counts”) or firing pattern (“spike timing”) of RA output was the best predictor of the behavioral group a given syllable would fall into... The final cost is the measure of dissimilarity, or “distance,” between the two trains... These distances are then used to independently classify the syllables they give rise to into behavioral groups, and those classifications are compared to the ones from the acoustic analysis... Next, the authors performed a second type of analysis, which examined the predictive information present in finer and finer timescale resolutions, from 40 milliseconds down to 1 millisecond, the shortest interval they could reliably measure... They found that as the temporal resolution increased, so did the information present, suggesting, in the authors' words, “that small differences in spike timing can significantly influence motor output. ” Indeed, the information present at that resolution was about ten times what was available from the spike rate alone... The results of these analyses have several implications... Most directly, restricting neural analysis to spike timing will likely significantly underestimate the information output of motor systems... Furthermore, while the spike trains analyzed here were from single neurons, the authors note that even higher-order information may be present in the coordinated output of multiple neurons... Finally, whether time-coded sensory inputs feed directly into time-coded motor outputs will require further investigation, but the demonstration of the existence of such outputs clears the way for those investigations. doi:10.1371/journal.pbio.1002018

Show MeSH