STDP Allows Close-to-Optimal Spatiotemporal Spike Pattern Detection by Single Coincidence Detector Neurons

Area of Science:

• Computational Neuroscience
• Neural Coding
• Synaptic Plasticity

Background:

• Repeating spatiotemporal spike patterns are known to encode information.
• The precise mechanisms by which downstream neurons extract this information remain largely unknown.

Purpose of the Study:

• To theoretically determine the optimal parameters for a single neuron to detect a specific spike pattern.
• To investigate if spike-timing-dependent plasticity (STDP) can enable neurons to achieve optimal pattern detection.

Main Methods:

• Analytical computation using a leaky integrate-and-fire (LIF) neuron model with homogeneous Poisson input.
• Simulation of a LIF neuron with additive STDP exposed to repeating input spike patterns.

Main Results:

• Optimal detection of spike patterns is typically achieved with a small membrane time constant (τ), even for long patterns.
• Spike-timing-dependent plasticity (STDP) can lead to optimal pattern detection without supervision.
• The optimal detector effectively acts as a coincidence detector, ignoring parts of the pattern due to fast memory decay.

Conclusions:

• Neurons may utilize fast membrane time constants and coincidence detection for efficient recognition of repeating sensory sequences.
• STDP provides a biological mechanism for unsupervised learning of temporal patterns.
• Coincidence detection may be a fundamental function of neurons in processing sequential information.