Limits...
Fast, scalable, Bayesian spike identification for multi-electrode arrays.

Prentice JS, Homann J, Simmons KD, Tkačik G, Balasubramanian V, Nelson PC - PLoS ONE (2011)

Bottom Line: Our method can distinguish large numbers of distinct neural units, even when spikes overlap, and accounts for intrinsic variability of spikes from each unit.Human interaction plays a key role in our method; but effort is minimized and streamlined via a graphical interface.We illustrate our method on data from guinea pig retinal ganglion cells and document its performance on simulated data consisting of spikes added to experimentally measured background noise.

View Article: PubMed Central - PubMed

Affiliation: Department of Physics and Astronomy, University of Pennsylvania, Philadelphia, Pennsylvania, United States of America. jprentic@sas.upenn.edu

ABSTRACT
We present an algorithm to identify individual neural spikes observed on high-density multi-electrode arrays (MEAs). Our method can distinguish large numbers of distinct neural units, even when spikes overlap, and accounts for intrinsic variability of spikes from each unit. As MEAs grow larger, it is important to find spike-identification methods that are scalable, that is, the computational cost of spike fitting should scale well with the number of units observed. Our algorithm accomplishes this goal, and is fast, because it exploits the spatial locality of each unit and the basic biophysics of extracellular signal propagation. Human interaction plays a key role in our method; but effort is minimized and streamlined via a graphical interface. We illustrate our method on data from guinea pig retinal ganglion cells and document its performance on simulated data consisting of spikes added to experimentally measured background noise. We present several tests demonstrating that the algorithm is highly accurate: it exhibits low error rates on fits to synthetic data, low refractory violation rates, good receptive field coverage, and consistency across users.

Show MeSH

Related in: MedlinePlus

Template building.(A) Detail of 40 of the aligned events used to compute a template, upsampled and shifted into alignment as described in the text. Some outlier traces reflect events in which this unit fired together with some other unit; the unwanted peaks occur at random times relative to the one of interest, and thus do not affect the template. (B) Blue, detail of a template waveform, showing the potential on 12 neighboring electrodes. Time in  runs horizontally; the vertical axis is potential in . Red, for comparison, the pointwise mean of the 430 waveforms used to find this template (nearly indistinguishable from the blue curve). (C) Detail of (A), showing only the leader channel. In addition, each trace has been rescaled by a constant to emphasize their similarity apart from variation in overall amplitude.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3140468&req=5

pone-0019884-g005: Template building.(A) Detail of 40 of the aligned events used to compute a template, upsampled and shifted into alignment as described in the text. Some outlier traces reflect events in which this unit fired together with some other unit; the unwanted peaks occur at random times relative to the one of interest, and thus do not affect the template. (B) Blue, detail of a template waveform, showing the potential on 12 neighboring electrodes. Time in runs horizontally; the vertical axis is potential in . Red, for comparison, the pointwise mean of the 430 waveforms used to find this template (nearly indistinguishable from the blue curve). (C) Detail of (A), showing only the leader channel. In addition, each trace has been rescaled by a constant to emphasize their similarity apart from variation in overall amplitude.

Mentions: Next we created a consensus waveform (“template”) summarizing each cluster of cropped, upsampled events, and characterized meaningful deviations from that consensus. We created a draft template by finding the pointwise median over all events in a cluster, then aligned each event to the draft template by maximizing their cross-correlation over time shifts, which we found to be more accurate than aligning to each event's peak time. Finally, we found the pointwise median (to suppress the effects of outliers) of the aligned events; this waveform was our template (Fig. 5B).


Fast, scalable, Bayesian spike identification for multi-electrode arrays.

Prentice JS, Homann J, Simmons KD, Tkačik G, Balasubramanian V, Nelson PC - PLoS ONE (2011)

Template building.(A) Detail of 40 of the aligned events used to compute a template, upsampled and shifted into alignment as described in the text. Some outlier traces reflect events in which this unit fired together with some other unit; the unwanted peaks occur at random times relative to the one of interest, and thus do not affect the template. (B) Blue, detail of a template waveform, showing the potential on 12 neighboring electrodes. Time in  runs horizontally; the vertical axis is potential in . Red, for comparison, the pointwise mean of the 430 waveforms used to find this template (nearly indistinguishable from the blue curve). (C) Detail of (A), showing only the leader channel. In addition, each trace has been rescaled by a constant to emphasize their similarity apart from variation in overall amplitude.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3140468&req=5

pone-0019884-g005: Template building.(A) Detail of 40 of the aligned events used to compute a template, upsampled and shifted into alignment as described in the text. Some outlier traces reflect events in which this unit fired together with some other unit; the unwanted peaks occur at random times relative to the one of interest, and thus do not affect the template. (B) Blue, detail of a template waveform, showing the potential on 12 neighboring electrodes. Time in runs horizontally; the vertical axis is potential in . Red, for comparison, the pointwise mean of the 430 waveforms used to find this template (nearly indistinguishable from the blue curve). (C) Detail of (A), showing only the leader channel. In addition, each trace has been rescaled by a constant to emphasize their similarity apart from variation in overall amplitude.
Mentions: Next we created a consensus waveform (“template”) summarizing each cluster of cropped, upsampled events, and characterized meaningful deviations from that consensus. We created a draft template by finding the pointwise median over all events in a cluster, then aligned each event to the draft template by maximizing their cross-correlation over time shifts, which we found to be more accurate than aligning to each event's peak time. Finally, we found the pointwise median (to suppress the effects of outliers) of the aligned events; this waveform was our template (Fig. 5B).

Bottom Line: Our method can distinguish large numbers of distinct neural units, even when spikes overlap, and accounts for intrinsic variability of spikes from each unit.Human interaction plays a key role in our method; but effort is minimized and streamlined via a graphical interface.We illustrate our method on data from guinea pig retinal ganglion cells and document its performance on simulated data consisting of spikes added to experimentally measured background noise.

View Article: PubMed Central - PubMed

Affiliation: Department of Physics and Astronomy, University of Pennsylvania, Philadelphia, Pennsylvania, United States of America. jprentic@sas.upenn.edu

ABSTRACT
We present an algorithm to identify individual neural spikes observed on high-density multi-electrode arrays (MEAs). Our method can distinguish large numbers of distinct neural units, even when spikes overlap, and accounts for intrinsic variability of spikes from each unit. As MEAs grow larger, it is important to find spike-identification methods that are scalable, that is, the computational cost of spike fitting should scale well with the number of units observed. Our algorithm accomplishes this goal, and is fast, because it exploits the spatial locality of each unit and the basic biophysics of extracellular signal propagation. Human interaction plays a key role in our method; but effort is minimized and streamlined via a graphical interface. We illustrate our method on data from guinea pig retinal ganglion cells and document its performance on simulated data consisting of spikes added to experimentally measured background noise. We present several tests demonstrating that the algorithm is highly accurate: it exhibits low error rates on fits to synthetic data, low refractory violation rates, good receptive field coverage, and consistency across users.

Show MeSH
Related in: MedlinePlus