Precise spiking motifs in neurobiological and neuromorphic data

Antoine Grimaldi, Amélie Gruel, Camille Besnainou, Jean-Nicolas Jérémie, Jean Martinet, Laurent U Perrinet
December 2022
Cite DOI Hal Code URL arXiv

Abstract

Why do neurons communicate through spikes? By definition, spikes are all-or-none neural events which occur at continuous times. In other words, spikes are on one side binary, existing or not without further details, and on the other can occur at any asynchronous time, without the need for a centralized clock. This stands in stark contrast to the analog representation of values and the discretized timing classically used in digital processing and at the base of modern-day neural networks. As neural systems almost systematically use this so-called event-based representation in the living world, a better understanding of this phenomenon remains a fundamental challenge in neurobiology in order to better interpret the profusion of recorded data. With the growing need for intelligent embedded systems, it also emerges as a new computing paradigm to enable the efficient operation of a new class of sensors and event-based computers, called neuromorphic, which could enable significant gains in computation time and energy consumption, a major societal issue in the era of the digital economy and global warming. In this review paper, we provide evidence from biology, theory and engineering that the precise timing of spikes plays a crucial role in our understanding of the efficiency of neural networks.

Type
Journal article
**Core mechanism of polychrony detection.** *(Left)* In this example, three presynaptic neurons denoted *b*, *c* and *d* are fully connected to two post-synaptic neurons *a* and *e*, with different delays of respectively 1, 5, and 9 ms for *a* and 8, 5, and 1 ms for *e*. *(Middle)* If three synchronous pulses are emitted from presynaptic neurons, this will generate post-synaptic potentials that will reach a and e asynchronously because of the heterogeneous delays, and they may not be sufficient to reach the membrane threshold in either of the post-synaptic neurons; therefore, no spike will be emitted, as this is not sufficient to reach the membrane threshold of the post synaptic neuron, so no output spike is emitted. *(Right)* If the pulses are emitted from presynaptic neurons such that, taking into account the delays, they reach the post-synaptic neuron *a* at the same time (here, at t = 10 ms), the post-synaptic potentials evoked by the three pre-synaptic neurons sum up, causing the voltage threshold to be crossed and thus to the emission of an output spike (red color), while none is emitted from post-synaptic neuron *e*.
Core mechanism of polychrony detection. (Left) In this example, three presynaptic neurons denoted b, c and d are fully connected to two post-synaptic neurons a and e, with different delays of respectively 1, 5, and 9 ms for a and 8, 5, and 1 ms for e. (Middle) If three synchronous pulses are emitted from presynaptic neurons, this will generate post-synaptic potentials that will reach a and e asynchronously because of the heterogeneous delays, and they may not be sufficient to reach the membrane threshold in either of the post-synaptic neurons; therefore, no spike will be emitted, as this is not sufficient to reach the membrane threshold of the post synaptic neuron, so no output spike is emitted. (Right) If the pulses are emitted from presynaptic neurons such that, taking into account the delays, they reach the post-synaptic neuron a at the same time (here, at t = 10 ms), the post-synaptic potentials evoked by the three pre-synaptic neurons sum up, causing the voltage threshold to be crossed and thus to the emission of an output spike (red color), while none is emitted from post-synaptic neuron e.
Antoine Grimaldi
Antoine Grimaldi
Phd candidate in Computational Neuroscience

During my PhD, I am focusing on Ultra-fast vision using Spiking Neural Networks.

Amélie Gruel
Amélie Gruel
PhD student in Computer Sciences at i3S/CNRS

During my PhD, I am focusing on Ultra-fast vision using Spiking Neural Networks.

Jean-Nicolas Jérémie
Jean-Nicolas Jérémie
Phd candidate in Computational Neuroscience

During my PhD, I am focusing on ultra-fast processing in event-based neural networks.

Jean Martinet
Jean Martinet
Laurent U Perrinet
Laurent U Perrinet
Researcher in Computational Neuroscience

My research interests include Machine Learning and computational neuroscience applied to Vision.