The Time Constant Rule for Neural Change Detection

Area of Science:

• Computational Neuroscience
• Sensory Biology
• Systems Neuroscience

Background:

• Sensory pathways exhibit diverse anatomy and function across species and modalities.
• A universal challenge is the rapid detection of meaningful environmental changes from noisy signals, crucial for survival.

Purpose of the Study:

• To propose a mechanism by which a single spiking neuron can achieve rapid change detection.
• To introduce and validate the 'time constant rule' linking neural voltage dynamics to online hypothesis testing.

Main Methods:

• Theoretical modeling of neural voltage dynamics and hypothesis testing.
• Analysis of published electrophysiological recordings from diverse sensory neurons (olfaction, mechanoreception, electroreception, audition) in insects and vertebrates.
• Comparison of measured membrane time constants with the predicted inverse relationship to the input rate change.

Main Results:

• The time constant rule, where a neuron's membrane time constant equals the inverse of its input rate change, was proposed.
• This rule demonstrates that subthreshold membrane potential represents a likelihood ratio of input rate change.
• Reanalysis of published data showed measured membrane time constants were consistent with the predicted rule across multiple sensory systems and species.

Conclusions:

• The time constant rule provides a direct interpretation of neural computation: voltage as an online likelihood ratio, time constant as statistical evidence decay, and spikes as change declarations.
• This rule bridges membrane biophysics, statistical computation, and the ecological function of neurons in rapid stimulus detection.
• The falsifiable nature of the rule invites further experimental investigation into its broader applicability across neural systems.