Limits...
Emergence of spatiotemporal invariance in large neuronal ensembles in rat barrel cortex.

Jacobs NS, Chen-Bee CH, Frostig RD - Front Neural Circuits (2015)

Bottom Line: Invariant sensory coding is the robust coding of some sensory information (e.g., stimulus type) despite major changes in other sensory parameters (e.g., stimulus strength).The contribution of large populations of neurons (ensembles) to invariant sensory coding is not well understood, but could offer distinct advantages over invariance in single cell receptive fields.We therefore suggest that ensemble-based invariance could provide a robust neurobiological substrate for invariant sensory coding and integration at an early stage of cortical sensory processing already in primary sensory cortex.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurobiology and Behavior, University of California, Irvine Irvine, CA, USA ; Department of Neurobiology and Behavior, Center for the Neurobiology of Learning and Memory, University of California, Irvine Irvine, CA, USA.

ABSTRACT
Invariant sensory coding is the robust coding of some sensory information (e.g., stimulus type) despite major changes in other sensory parameters (e.g., stimulus strength). The contribution of large populations of neurons (ensembles) to invariant sensory coding is not well understood, but could offer distinct advantages over invariance in single cell receptive fields. To test invariant sensory coding in neuronal ensembles evoked by single whisker stimulation as early as primary sensory cortex, we recorded detailed spatiotemporal movies of evoked ensemble activity through the depth of rat barrel cortex using microelectrode arrays. We found that an emergent property of whisker evoked ensemble activity, its spatiotemporal profile, was notably invariant across major changes in stimulus amplitude (up to >200-fold). Such ensemble-based invariance was found for single whisker stimulation as well as for the integrated profile of activity evoked by the more naturalistic stimulation of the entire whisker array. Further, the integrated profile of whisker array evoked ensemble activity and its invariance to stimulus amplitude shares striking similarities to "funneled" tactile perception in humans. We therefore suggest that ensemble-based invariance could provide a robust neurobiological substrate for invariant sensory coding and integration at an early stage of cortical sensory processing already in primary sensory cortex.

No MeSH data available.


Whisker array evoked LFP for the first deflection. (A,B) Raw (A) and peak-normalized (B) movies of averaged (n = 7) whisker array evoked LFP for the first deflection. (C,D) Continuous quantification of spatial profiles with PCA loadings (D) and similarity between spatial profiles with correlations (E). Traces are mean ± s.e.m. Gray shaded regions indicate mean onset and peak latencies (± s.e.m). (E) Raw and normalized mean onset frames. (F–H) Raw and normalized mean peak frames (F), broken down by layer (G), and quantification of peak response properties (H). *p < 0.05.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4495341&req=5

Figure 5: Whisker array evoked LFP for the first deflection. (A,B) Raw (A) and peak-normalized (B) movies of averaged (n = 7) whisker array evoked LFP for the first deflection. (C,D) Continuous quantification of spatial profiles with PCA loadings (D) and similarity between spatial profiles with correlations (E). Traces are mean ± s.e.m. Gray shaded regions indicate mean onset and peak latencies (± s.e.m). (E) Raw and normalized mean onset frames. (F–H) Raw and normalized mean peak frames (F), broken down by layer (G), and quantification of peak response properties (H). *p < 0.05.

Mentions: Further supporting the similarity of spatial profiles, peak frames were also highly correlated with each other (mean r2 = 0.95 ± < 0.01; grand mean of comparison and subject; Figure 3H, right, closed circles). A simple internal control was used to test the sensitivity of correlations by comparing single whisker to whisker array responses, which are both characterized by the same basic shape (a single, central peak of activity; see schematics in Figure 2B). Importantly, this control comparison resulted in significantly lower r2-values [Figure 3H, right, open circles; grand mean r2 = 0.62 ± 0.06; F(1, 5) = 107.86, p < 0.001; for peak frames of whisker array evoked LFP see Figure 5F], indicating that correlations were highly sensitive to even subtle changes in the profile of evoked activity.


Emergence of spatiotemporal invariance in large neuronal ensembles in rat barrel cortex.

Jacobs NS, Chen-Bee CH, Frostig RD - Front Neural Circuits (2015)

Whisker array evoked LFP for the first deflection. (A,B) Raw (A) and peak-normalized (B) movies of averaged (n = 7) whisker array evoked LFP for the first deflection. (C,D) Continuous quantification of spatial profiles with PCA loadings (D) and similarity between spatial profiles with correlations (E). Traces are mean ± s.e.m. Gray shaded regions indicate mean onset and peak latencies (± s.e.m). (E) Raw and normalized mean onset frames. (F–H) Raw and normalized mean peak frames (F), broken down by layer (G), and quantification of peak response properties (H). *p < 0.05.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4495341&req=5

Figure 5: Whisker array evoked LFP for the first deflection. (A,B) Raw (A) and peak-normalized (B) movies of averaged (n = 7) whisker array evoked LFP for the first deflection. (C,D) Continuous quantification of spatial profiles with PCA loadings (D) and similarity between spatial profiles with correlations (E). Traces are mean ± s.e.m. Gray shaded regions indicate mean onset and peak latencies (± s.e.m). (E) Raw and normalized mean onset frames. (F–H) Raw and normalized mean peak frames (F), broken down by layer (G), and quantification of peak response properties (H). *p < 0.05.
Mentions: Further supporting the similarity of spatial profiles, peak frames were also highly correlated with each other (mean r2 = 0.95 ± < 0.01; grand mean of comparison and subject; Figure 3H, right, closed circles). A simple internal control was used to test the sensitivity of correlations by comparing single whisker to whisker array responses, which are both characterized by the same basic shape (a single, central peak of activity; see schematics in Figure 2B). Importantly, this control comparison resulted in significantly lower r2-values [Figure 3H, right, open circles; grand mean r2 = 0.62 ± 0.06; F(1, 5) = 107.86, p < 0.001; for peak frames of whisker array evoked LFP see Figure 5F], indicating that correlations were highly sensitive to even subtle changes in the profile of evoked activity.

Bottom Line: Invariant sensory coding is the robust coding of some sensory information (e.g., stimulus type) despite major changes in other sensory parameters (e.g., stimulus strength).The contribution of large populations of neurons (ensembles) to invariant sensory coding is not well understood, but could offer distinct advantages over invariance in single cell receptive fields.We therefore suggest that ensemble-based invariance could provide a robust neurobiological substrate for invariant sensory coding and integration at an early stage of cortical sensory processing already in primary sensory cortex.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurobiology and Behavior, University of California, Irvine Irvine, CA, USA ; Department of Neurobiology and Behavior, Center for the Neurobiology of Learning and Memory, University of California, Irvine Irvine, CA, USA.

ABSTRACT
Invariant sensory coding is the robust coding of some sensory information (e.g., stimulus type) despite major changes in other sensory parameters (e.g., stimulus strength). The contribution of large populations of neurons (ensembles) to invariant sensory coding is not well understood, but could offer distinct advantages over invariance in single cell receptive fields. To test invariant sensory coding in neuronal ensembles evoked by single whisker stimulation as early as primary sensory cortex, we recorded detailed spatiotemporal movies of evoked ensemble activity through the depth of rat barrel cortex using microelectrode arrays. We found that an emergent property of whisker evoked ensemble activity, its spatiotemporal profile, was notably invariant across major changes in stimulus amplitude (up to >200-fold). Such ensemble-based invariance was found for single whisker stimulation as well as for the integrated profile of activity evoked by the more naturalistic stimulation of the entire whisker array. Further, the integrated profile of whisker array evoked ensemble activity and its invariance to stimulus amplitude shares striking similarities to "funneled" tactile perception in humans. We therefore suggest that ensemble-based invariance could provide a robust neurobiological substrate for invariant sensory coding and integration at an early stage of cortical sensory processing already in primary sensory cortex.

No MeSH data available.