Local Field Potential

Local Field Potential (LFP) is the electric potential recorded in the extracellular space near a population of neurons. It reflects the summed synaptic currents, action potentials, and glial currents within a local volume of brain tissue — typically within a few hundred micrometers of the recording electrode. Unlike the EEG, which is recorded at the scalp and reflects the activity of large cortical areas, the LFP is a mesoscale signal that captures the dynamics of local neural circuits with spatial and temporal precision that bridges the gap between single-neuron spikes and global brain rhythms.

The LFP is the primary signal recorded by invasive electrodes in animal research and by clinical depth electrodes in human patients. It is the substrate from which the scalp EEG is ultimately derived, filtered and spatially smeared by the skull and scalp. Understanding the LFP is therefore essential for interpreting EEG and for understanding how local neural synchrony becomes global coherence.

What the LFP Measures

The LFP is not a single, well-defined physiological quantity. It is the voltage difference between two points in the extracellular medium, and its composition depends on the electrode's position, the geometry of the local tissue, and the frequency band of interest. At low frequencies (below ~100 Hz), the LFP is dominated by synaptic currents — the currents that flow into and out of neurons as they receive excitatory and inhibitory inputs. These currents create potential gradients in the extracellular space that are detectable as LFP fluctuations.

At higher frequencies, the LFP includes contributions from action potentials, which are brief, high-amplitude events that can be detected extracellularly as spikes. However, the relationship between spikes and the LFP is complex. A single spike produces a tiny extracellular potential; the LFP reflects the coherent firing of many neurons, not individual action potentials. The separation of spikes and LFP — often done by high-pass filtering for spikes and low-pass filtering for LFP — is a methodological convenience, not a fundamental physiological distinction.

The LFP also reflects the activity of glial cells, which buffer extracellular potassium and produce slow potential shifts. In addition, ionic diffusion, capacitive currents, and ephaptic coupling (direct electrical interaction between neurons through the extracellular medium) contribute to the LFP signal. The LFP is, in short, a composite signal that encodes the collective electrical activity of a neural microcircuit.

Frequency Content and Cognitive Correlates

Like the EEG, the LFP is typically analyzed in the frequency domain, where its power spectrum reveals characteristic rhythmic activity. The LFP captures oscillations that are often sharper and more localized than their scalp EEG counterparts:

• Theta rhythms (4–8 Hz) are prominent in the hippocampus and entorhinal cortex, where they organize the temporal structure of spatial navigation and memory encoding. The theta rhythm is not a single oscillator but a traveling wave that propagates through the hippocampus, phase-locking the firing of pyramidal cells and interneurons to specific phases of the cycle.
• Gamma rhythms (30–100 Hz) are ubiquitous in cortex and hippocampus and are thought to underlie the binding of distributed neural representations into coherent percepts. Gamma oscillations in the LFP are often nested within theta cycles, a phenomenon known as theta-gamma coupling that may provide a multi-timescale code for information organization.
• Sharp-wave ripples (100–250 Hz) are brief, high-frequency events in the hippocampal LFP that occur during sleep and quiet wakefulness. They are associated with the replay of previously experienced sequences of neural activity and are believed to be essential for memory consolidation.

These LFP rhythms are not epiphenomena. They are causally involved in neural computation. Pharmacological suppression of theta or gamma oscillations impairs memory and perception, and artificial induction of oscillations can enhance or disrupt cognitive performance. The LFP is not merely a readout of neural activity; it is a participant in the dynamics it measures.

Current Source Density and Spatial Localization

The LFP is a volume-conducted signal, and its spatial distribution depends on the geometry of the current sources and sinks in the tissue. Current Source Density (CSD) analysis is a mathematical technique that estimates the spatial distribution of transmembrane currents from multiple LFP recordings arranged in a linear or planar array. By taking the second spatial derivative of the LFP, CSD analysis localizes the current sources and sinks, providing a sharper picture of where synaptic activity is concentrated.

CSD analysis is particularly powerful when combined with laminar recordings in cortex, where electrodes span the cortical layers from the superficial layers to the deep layers. Each layer has distinct patterns of inputs and outputs, and the laminar CSD profile reveals the flow of information through the cortical column. This laminar analysis is the closest non-invasive equivalent to understanding the cortical circuit's internal dynamics.

LFP as a Systems-Level Phenomenon

The LFP is a paradigmatic example of an emergent, systems-level signal. It is not produced by any single neuron or synapse but by the coordinated activity of thousands of cells. The individual synaptic current is microscopic and transient; the LFP is macroscopic and sustained. The relationship between the microscopic currents and the macroscopic field is a problem in nonequilibrium statistical mechanics: how do the stochastic dynamics of individual neurons produce coherent, rhythmic population activity?

This question has been addressed through mean-field models and neural mass models, which describe the average activity of neural populations rather than individual neurons. These models treat the LFP as a proxy for the population firing rate and use it to fit parameters that describe the strength and time constants of excitatory and inhibitory interactions. The resulting models can predict the LFP power spectrum and its response to perturbations, but they do not fully capture the biophysical complexity of the LFP signal.

The LFP's most important conceptual role is as a bridge between scales. It connects the single-neuron spike trains recorded by electrophysiologists to the global brain rhythms measured by EEG and MEG. It connects the synaptic currents described by the Hodgkin-Huxley model to the network dynamics described by neural mass models. And it connects the cellular physiology of individual neurons to the cognitive states studied by systems neuroscience. The LFP is the Rosetta Stone of neuroscience — a signal that can be read, with appropriate translation, at multiple levels of description.

_The LFP is not a blurry version of the spike; it is a different signal entirely, carrying information that spikes alone cannot provide. The temporal pattern of synaptic currents, the spatial organization of population activity, the phase relationships between different frequencies — all of these are encoded in the LFP and are invisible to spike trains. The insistence that spikes are the real neural signal and the LFP is merely an epiphenomenon is a prejudice of the reductionist program, not a conclusion from the data. The brain speaks in many channels simultaneously, and the LFP is one of the most important._