A model of cortical timbre classification using spike latency coding

2021 ◽  
Vol 149 (4) ◽  
pp. A77-A77
Author(s):  
David A. Dahlbom ◽  
Jonas Braasch
Keyword(s):  
Spike Latency ◽  
Latency Coding

Bat echolocation processing using first-spike latency coding

2009 ◽  
Vol 22 (10) ◽  
pp. 1372-1382 ◽  
Author(s):  
Bertrand Fontaine ◽  
Herbert Peremans
Keyword(s):  
Spike Latency ◽  
Latency Coding

Spike Latency Coding in Biologically Inspired Microelectronic Nose

2011 ◽  
Vol 5 (2) ◽  
pp. 160-168 ◽  
Author(s):  
Hung Tat Chen ◽  
Kwan Ting Ng ◽  
Amine Bermak ◽  
Man Kay Law ◽  
Dominique Martinez
Keyword(s):  
Biologically Inspired ◽  
Spike Latency ◽  
Latency Coding

Dopamine modulation of two subthreshold currents produces phase shifts in activity of an identified motoneuron

1995 ◽  
Vol 74 (4) ◽  
pp. 1404-1420 ◽  
Author(s):  
R. M. Harris-Warrick ◽  
L. M. Coniglio ◽  
R. M. Levini ◽  
S. Gueron ◽  
J. Guckenheimer
Keyword(s):  
Action Potentials ◽  
Voltage Dependence ◽  
Motor Pattern ◽  
Outward Current ◽  
Untreated Cell ◽  
Slight Reduction ◽  
Inward Current ◽  
Postinhibitory Rebound ◽  
Spike Latency ◽  
Dopamine Modulation

1. The lateral pyloric (LP) neuron is a component of the 14-neuron pyloric central pattern generator in the stomatogastric ganglion of the spiny lobster, Panulirus interruptus. In the pyloric rhythm, this neuron fires rhythmic bursts of action potentials whose phasing depends on the pattern of synaptic inhibition from other network neurons and on the intrinsic postinhibitory rebound properties of the LP cell itself. Bath-applied dopamine excites the LP cell and causes its activity to be phase advanced in the pyloric motor pattern. At least part of this modulatory effect is due to dopaminergic modulation of the intrinsic rate of postinhibitory rebound in the LP cell. 2. The LP neuron was isolated from all detectable synaptic input. We measured the rate of recovery after 1-s hyperpolarizing current injections of varying amplitudes, quantifying the latency to the first spike following the hyperpolarizing prepulse and the interval between the first and second action potentials. Dopamine reduced both the first spike latency and the first interspike interval (ISI) in the isolated LP neuron. During the hyperpolarizating pre-steps, the LP cell showed a slow depolarizing sag voltage that was enhanced by dopamine. 3. We used voltage clamp to analyze dopamine modulation of subthreshold ionic currents whose activity is affected by hyperpolarizing prepulses. Dopamine modulated the transient potassium current IA by reducing its maximal conductance and shifting its voltage dependence for activation and inactivation to more depolarized voltages. This outward current is normally transiently activated after hyperpolarization of the LP cell, and delays the rate of postinhibitory rebound; by reducing IA, dopamine thus accelerates the rate of rebound of the LP neuron. 4. Dopamine also modulated the hyperpolarization-activated inward current Ih by shifting its voltage dependence for activation 20 mV in the depolarizing direction and accelerating its rate of activation. This enhanced inward current helps accelerate the rate of rebound in the LP cell after inhibition. 5. The relative roles of Ih and IA in determining the first spike latency and first ISI were explored using pharmacological blockers of Ih (Cs+) and IA [4-aminopyridine (4-AP)]. Blockade of Ih prolonged the first spike latency and first ISI, but only slightly reduced the net effect of dopamine. In the continued presence of Cs+, blockade of IA with 4-AP greatly shortened the first spike latency and first ISI. Under conditions where both Ih and IA were blocked, dopamine had no additional effect on the LP cell. 6. We used the dynamic clamp technique to further study the relative roles of IA and Ih modulation in dopamine's phase advance of the LP cell. We blocked the endogenous Ih with Cs+ and replaced it with a simulated current generated by a computer model of Ih. The neuron with simulated Ih gave curves relating the hyperpolarizing prepulse amplitude to first spike latency that were the same as in the untreated cell. Changing the computer parameters of the simulated Ih to those induced by dopamine without changing IA caused only a slight reduction in first spike latency, which was approximately 20% of the total reduction caused by dopamine in an untreated cell. Bath application of dopamine in the presence of Cs+ and simulated Ih (with control parameters) allowed us to determine the effect of altering IA but not Ih: this caused a significant reduction in first spike latency, but it was still only approximately 70% of the effect of dopamine in the untreated cell. Finally, in the continued presence of dopamine, changing the parameters of the simulated Ih to those observed with dopamine reduced the first spike latency to that seen with dopamine in the untreated cell. 7. We generated a mathematical model of the lobster LP neuron, based on the model of Buchholtz et al. for the crab LP neuron.

Spike-latency codes and the effect of saccades

2005 ◽  
Vol 65-66 ◽  
pp. 189-194 ◽  
Author(s):  
Rüdiger Kupper ◽  
Marc-Oliver Gewaltig ◽  
Ursula Körner ◽  
Edgar Körner
Keyword(s):  
Spike Latency

scholarly journals nMNSD—A Spiking Neuron-Based Classifier That Combines Weight-Adjustment and Delay-Shift

2021 ◽  
Vol 15 ◽  
Author(s):  
Gianluca Susi ◽  
Luis F. Antón-Toro ◽  
Fernando Maestú ◽  
Ernesto Pereda ◽  
Claudio Mirasso
Keyword(s):  
State Of The Art ◽  
Classification Problem ◽  
Mathematical Framework ◽  
Recognition Mechanism ◽  
Specific Input ◽  
Biological Learning ◽  
Time And Energy ◽  
Spike Latency ◽  
Spike Sequence

The recent “multi-neuronal spike sequence detector” (MNSD) architecture integrates the weight- and delay-adjustment methods by combining heterosynaptic plasticity with the neurocomputational feature spike latency, representing a new opportunity to understand the mechanisms underlying biological learning. Unfortunately, the range of problems to which this topology can be applied is limited because of the low cardinality of the parallel spike trains that it can process, and the lack of a visualization mechanism to understand its internal operation. We present here the nMNSD structure, which is a generalization of the MNSD to any number of inputs. The mathematical framework of the structure is introduced, together with the “trapezoid method,” that is a reduced method to analyze the recognition mechanism operated by the nMNSD in response to a specific input parallel spike train. We apply the nMNSD to a classification problem previously faced with the classical MNSD from the same authors, showing the new possibilities the nMNSD opens, with associated improvement in classification performances. Finally, we benchmark the nMNSD on the classification of static inputs (MNIST database) obtaining state-of-the-art accuracies together with advantageous aspects in terms of time- and energy-efficiency if compared to similar classification methods.

scholarly journals Low error discrimination using a correlated population code

2012 ◽  
Vol 108 (4) ◽  
pp. 1069-1088 ◽  
Author(s):  
Greg Schwartz ◽  
Jakob Macke ◽  
Dario Amodei ◽  
Hanlin Tang ◽  
Michael J. Berry
Keyword(s):  
Spatial Information ◽  
Ganglion Cells ◽  
Discrimination Performance ◽  
Spike Timing ◽  
Multiple Time ◽  
Entropy Model ◽  
Decoding Algorithms ◽  
Single Spike ◽  
Order Of Magnitude ◽  
Spike Latency

We explored the manner in which spatial information is encoded by retinal ganglion cell populations. We flashed a set of 36 shape stimuli onto the tiger salamander retina and used different decoding algorithms to read out information from a population of 162 ganglion cells. We compared the discrimination performance of linear decoders, which ignore correlation induced by common stimulation, with nonlinear decoders, which can accurately model these correlations. Similar to previous studies, decoders that ignored correlation suffered only a modest drop in discrimination performance for groups of up to ∼30 cells. However, for more realistic groups of 100+ cells, we found order-of-magnitude differences in the error rate. We also compared decoders that used only the presence of a single spike from each cell with more complex decoders that included information from multiple spike counts and multiple time bins. More complex decoders substantially outperformed simpler decoders, showing the importance of spike timing information. Particularly effective was the first spike latency representation, which allowed zero discrimination errors for the majority of shape stimuli. Furthermore, the performance of nonlinear decoders showed even greater enhancement compared with linear decoders for these complex representations. Finally, decoders that approximated the correlation structure in the population by matching all pairwise correlations with a maximum entropy model fit to all 162 neurons were quite successful, especially for the spike latency representation. Together, these results suggest a picture in which linear decoders allow a coarse categorization of shape stimuli, whereas nonlinear decoders, which take advantage of both correlation and spike timing, are needed to achieve high-fidelity discrimination.

Sign in / Sign up

Export Citation Format

Share Document