Neural excitability

Neural excitability is the dynamical property of a neuron or neural population that determines its response to input: whether a stimulus produces no output (subthreshold response), a single action potential, or sustained periodic firing. It is not a static threshold but a dynamical regime — a region of parameter space in which the system's phase portrait contains the geometric structures that support threshold behavior, rebound firing, and burst generation.

The mathematical classification of neural excitability, developed by Eugene Izhikevich and others, identifies two fundamental types based on the bifurcation mechanism by which a resting neuron loses stability and begins to fire. Type I excitability arises through a saddle-node bifurcation on a limit cycle: the neuron transitions from quiescence to firing with arbitrarily low frequency, and the firing rate increases continuously with input strength. Type II excitability arises through a Hopf bifurcation: the neuron begins firing at a finite, nonzero frequency, and the firing rate jumps discontinuously at threshold.

This bifurcation-based classification is not merely taxonomic. It predicts distinct neural behaviors. Type I neurons, common in cortical pyramidal cells, integrate inputs smoothly and fire at rates proportional to stimulus intensity — they are integrators. Type II neurons, common in fast-spiking interneurons, fire at a preferred frequency and are less sensitive to gradual input changes — they are resonators. The distinction has implications for neural coding, synchronization, and the design of brain-computer interfaces.

Beyond the type I/II dichotomy, neural excitability encompasses more complex regimes: burst excitability (where brief inputs trigger sustained volleys of spikes), rebound excitability (where inhibition releases a post-inhibitory burst), and canard-mediated transitions between subthreshold oscillation and full action potentials. Each regime corresponds to a different geometry in the neuron's phase space and a different bifurcation structure in its parameter space.

From a systems perspective, neural excitability is the mechanism by which a biological system converts continuous analog input into discrete digital output — the fundamental information-processing primitive of the nervous system. The excitability threshold is not a passive barrier but an active dynamical structure, shaped by ion channel densities, membrane capacitance, and synaptic connectivity, that determines how the nervous system maps the continuous world into discrete events.