Limits...
Neuronal precision and the limits for acoustic signal recognition in a small neuronal network.

Neuhofer D, Stemmler M, Ronacher B - J. Comp. Physiol. A Neuroethol. Sens. Neural. Behav. Physiol. (2010)

Bottom Line: By progressively corrupting the envelope of a female song, we determined the critical degradation level at which males failed to recognize a courtship call in behavioral experiments.At consecutive levels of processing, intrinsic variability increased, while the sensitivity to external noise decreased.We followed two approaches to determine critical degradation levels from spike train dissimilarities, and compared the results with the limits of signal recognition measured in behaving animals.

View Article: PubMed Central - PubMed

Affiliation: Department of Biology, Humboldt-Universität zu Berlin, Invalidenstrasse 43, 10115, Berlin, Germany. neuhofda@cms.hu-berlin.de

ABSTRACT
Recognition of acoustic signals may be impeded by two factors: extrinsic noise, which degrades sounds before they arrive at the receiver's ears, and intrinsic neuronal noise, which reveals itself in the trial-to-trial variability of the responses to identical sounds. Here we analyzed how these two noise sources affect the recognition of acoustic signals from potential mates in grasshoppers. By progressively corrupting the envelope of a female song, we determined the critical degradation level at which males failed to recognize a courtship call in behavioral experiments. Using the same stimuli, we recorded intracellularly from auditory neurons at three different processing levels, and quantified the corresponding changes in spike train patterns by a spike train metric, which assigns a distance between spike trains. Unexpectedly, for most neurons, intrinsic variability accounted for the main part of the metric distance between spike trains, even at the strongest degradation levels. At consecutive levels of processing, intrinsic variability increased, while the sensitivity to external noise decreased. We followed two approaches to determine critical degradation levels from spike train dissimilarities, and compared the results with the limits of signal recognition measured in behaving animals.

Show MeSH

Related in: MedlinePlus

a Linear reconstruction of the uncorrupted song envelope (solid lines) or the noisy stimulus envelope (dashed lines). Reconstruction performance was quantified as the fraction of envelope variance explained by the reconstruction (coding fraction). Four cell types are illustrated with different colors and symbols (see inset). b The optimal reconstruction filters depended on the neuron. Red lines indicate the filter for the original song envelope, while the green lines represent the filter for reconstructing the noisy envelope for a degradation level of 6 dB (shown in green). c The intrinsic variability (abscissa) was inversely correlated to the coding fraction (ordinate). Linear regression lines for the different computation levels and the two species are shown in different colors
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3040818&req=5

Fig7: a Linear reconstruction of the uncorrupted song envelope (solid lines) or the noisy stimulus envelope (dashed lines). Reconstruction performance was quantified as the fraction of envelope variance explained by the reconstruction (coding fraction). Four cell types are illustrated with different colors and symbols (see inset). b The optimal reconstruction filters depended on the neuron. Red lines indicate the filter for the original song envelope, while the green lines represent the filter for reconstructing the noisy envelope for a degradation level of 6 dB (shown in green). c The intrinsic variability (abscissa) was inversely correlated to the coding fraction (ordinate). Linear regression lines for the different computation levels and the two species are shown in different colors

Mentions: As the spike train pattern becomes corrupted by noise, it should become more difficult to reconstruct the song from the spike trains. This was indeed the case. Figure 7a illustrates the coding fraction for four different cell types as a function of the noise level, where we compare the reconstruction of each degraded envelope with the attempt to reconstruct the original signal by itself. Noise changes the envelope significantly (Fig. 1a), progressively masking the original signal’s periodic structure and leading to lower coding fractions. The repetition of rhythmic elements in the stimulus facilitates reconstruction, which is why reconstructing the original stimulus in the presence of noise is as successful as reconstructing the true, noisy envelope, irrespective of the degradation level (compare stippled vs. continuous lines in Fig. 7a). The optimal linear reconstruction filters for the original and the corrupted envelope signal, while highly different across cell types, change little as the noise degradation level increases, as shown in Fig. 7b. Receptors and TN1 cells respond to each sound pulse in a syllable, which leads to a single, large positive peak in the filter before the spike occurred (dashed line). The AN12 responds to syllable onsets, so that the filter’s shape mirrors a single syllable. Correlating coding fractions with spike train distances in Fig. 7c reveals the impact of intrinsic variability on the coding fraction of the uncorrupted stimulus. At all processing levels reconstruction quality correlates negatively with intrinsic distance (REC: r = −0.575; n.s. (L.m.); LN: r = −0.873; p < 0.001 (C.b.), r = −0.821; p < 0.001 (L.m.); AN: r = −0.811; p < 0.001 (C.b.), r = −0.469; p > 0.05 (L.m.); Pearson’s correlation). Receptor cells exhibited significantly higher coding fractions (median = 0.78) than local interneurons (median = 0.66) and ascending neurons (median = 0.31) (REC–LN n.s., LN–AN: p < 0.001). These results demonstrate that the high intrinsic variability of many ascending neurons has a dramatic effect on the amount of information that can be linearly decoded from the spike trains.Fig. 7


Neuronal precision and the limits for acoustic signal recognition in a small neuronal network.

Neuhofer D, Stemmler M, Ronacher B - J. Comp. Physiol. A Neuroethol. Sens. Neural. Behav. Physiol. (2010)

a Linear reconstruction of the uncorrupted song envelope (solid lines) or the noisy stimulus envelope (dashed lines). Reconstruction performance was quantified as the fraction of envelope variance explained by the reconstruction (coding fraction). Four cell types are illustrated with different colors and symbols (see inset). b The optimal reconstruction filters depended on the neuron. Red lines indicate the filter for the original song envelope, while the green lines represent the filter for reconstructing the noisy envelope for a degradation level of 6 dB (shown in green). c The intrinsic variability (abscissa) was inversely correlated to the coding fraction (ordinate). Linear regression lines for the different computation levels and the two species are shown in different colors
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3040818&req=5

Fig7: a Linear reconstruction of the uncorrupted song envelope (solid lines) or the noisy stimulus envelope (dashed lines). Reconstruction performance was quantified as the fraction of envelope variance explained by the reconstruction (coding fraction). Four cell types are illustrated with different colors and symbols (see inset). b The optimal reconstruction filters depended on the neuron. Red lines indicate the filter for the original song envelope, while the green lines represent the filter for reconstructing the noisy envelope for a degradation level of 6 dB (shown in green). c The intrinsic variability (abscissa) was inversely correlated to the coding fraction (ordinate). Linear regression lines for the different computation levels and the two species are shown in different colors
Mentions: As the spike train pattern becomes corrupted by noise, it should become more difficult to reconstruct the song from the spike trains. This was indeed the case. Figure 7a illustrates the coding fraction for four different cell types as a function of the noise level, where we compare the reconstruction of each degraded envelope with the attempt to reconstruct the original signal by itself. Noise changes the envelope significantly (Fig. 1a), progressively masking the original signal’s periodic structure and leading to lower coding fractions. The repetition of rhythmic elements in the stimulus facilitates reconstruction, which is why reconstructing the original stimulus in the presence of noise is as successful as reconstructing the true, noisy envelope, irrespective of the degradation level (compare stippled vs. continuous lines in Fig. 7a). The optimal linear reconstruction filters for the original and the corrupted envelope signal, while highly different across cell types, change little as the noise degradation level increases, as shown in Fig. 7b. Receptors and TN1 cells respond to each sound pulse in a syllable, which leads to a single, large positive peak in the filter before the spike occurred (dashed line). The AN12 responds to syllable onsets, so that the filter’s shape mirrors a single syllable. Correlating coding fractions with spike train distances in Fig. 7c reveals the impact of intrinsic variability on the coding fraction of the uncorrupted stimulus. At all processing levels reconstruction quality correlates negatively with intrinsic distance (REC: r = −0.575; n.s. (L.m.); LN: r = −0.873; p < 0.001 (C.b.), r = −0.821; p < 0.001 (L.m.); AN: r = −0.811; p < 0.001 (C.b.), r = −0.469; p > 0.05 (L.m.); Pearson’s correlation). Receptor cells exhibited significantly higher coding fractions (median = 0.78) than local interneurons (median = 0.66) and ascending neurons (median = 0.31) (REC–LN n.s., LN–AN: p < 0.001). These results demonstrate that the high intrinsic variability of many ascending neurons has a dramatic effect on the amount of information that can be linearly decoded from the spike trains.Fig. 7

Bottom Line: By progressively corrupting the envelope of a female song, we determined the critical degradation level at which males failed to recognize a courtship call in behavioral experiments.At consecutive levels of processing, intrinsic variability increased, while the sensitivity to external noise decreased.We followed two approaches to determine critical degradation levels from spike train dissimilarities, and compared the results with the limits of signal recognition measured in behaving animals.

View Article: PubMed Central - PubMed

Affiliation: Department of Biology, Humboldt-Universität zu Berlin, Invalidenstrasse 43, 10115, Berlin, Germany. neuhofda@cms.hu-berlin.de

ABSTRACT
Recognition of acoustic signals may be impeded by two factors: extrinsic noise, which degrades sounds before they arrive at the receiver's ears, and intrinsic neuronal noise, which reveals itself in the trial-to-trial variability of the responses to identical sounds. Here we analyzed how these two noise sources affect the recognition of acoustic signals from potential mates in grasshoppers. By progressively corrupting the envelope of a female song, we determined the critical degradation level at which males failed to recognize a courtship call in behavioral experiments. Using the same stimuli, we recorded intracellularly from auditory neurons at three different processing levels, and quantified the corresponding changes in spike train patterns by a spike train metric, which assigns a distance between spike trains. Unexpectedly, for most neurons, intrinsic variability accounted for the main part of the metric distance between spike trains, even at the strongest degradation levels. At consecutive levels of processing, intrinsic variability increased, while the sensitivity to external noise decreased. We followed two approaches to determine critical degradation levels from spike train dissimilarities, and compared the results with the limits of signal recognition measured in behaving animals.

Show MeSH
Related in: MedlinePlus