EEG and MEG: Reading Your Brain's Electrical and Magnetic Signals

In 1929, German psychiatrist Hans Berger attached silver electrodes to a patient's scalp and recorded something no one had ever seen: the brain's own electrical rhythms, captured while the person sat quietly with eyes closed. He called his device the Elektroenkephalogramm. Nearly a century later, neuroscientists can capture brain signals with sub-millisecond timing, revealing exactly when the visual cortex responds to a flash of light (around 100 ms) or when language areas activate as you read a sentence (around 400 ms). EEG and its magnetic cousin MEG are the tools that make this precision possible.

By the end of this lesson you will be able to: - Explain how neurons generate the electrical and magnetic signals that EEG and MEG detect - Describe the hardware, setup, and recording principles of both techniques - Distinguish the five major brainwave frequency bands and their behavioral correlates - Compare the spatial and temporal resolution of EEG versus MEG - Design a simple event-related potential (ERP) experiment

The brain contains roughly 86 billion neurons. When a neuron fires, it produces a brief action potential — a spike of electrical activity lasting about 1 millisecond. However, individual spikes are too small and too scattered to detect from outside the skull. EEG and MEG actually detect a different signal: excitatory post-synaptic potentials (EPSPs), which are slower electrical currents flowing along the dendrites of pyramidal neurons in the cortex. When thousands of these neurons fire in synchrony and are oriented perpendicular to the scalp, their combined electrical fields sum up enough to be captured by scalp electrodes. Key physics: a moving electrical charge creates a magnetic field (Ampere's law). So the same synchronized dendritic current that EEG captures electrically, MEG detects magnetically — both techniques are measuring the same underlying neural event from different physical perspectives.

EEG records voltage fluctuations at the scalp using an array of electrodes — typically 32, 64, or 256 channels — placed according to the international 10-20 system (named for the 10% and 20% intervals between electrode positions relative to skull landmarks). Brainwave frequency bands revealed by EEG: Delta (0.5-4 Hz): deep slow-wave sleep; dominant in infants. Theta (4-8 Hz): drowsiness and memory encoding; peaks in hippocampus during navigation tasks. Alpha (8-13 Hz): relaxed wakefulness with eyes closed; Berger's original discovery in 1929. Beta (13-30 Hz): active thinking and focused attention; suppressed during movement. Gamma (30-100 Hz): high-level cognitive binding and feature integration; very difficult to measure cleanly at the scalp. EEG's great strength is its temporal resolution: it tracks brain changes at the scale of milliseconds. Its weakness is spatial resolution — because the signal travels through cerebrospinal fluid, skull, and scalp, pinpointing exactly which brain region generated it is difficult (typically 1-2 cm accuracy at best).

MEG detects the tiny magnetic fields produced by the same dendritic currents that EEG records electrically. These magnetic fields are extraordinarily weak — roughly 100 femtotesla (fT), or about one billion times weaker than Earth's magnetic field. To detect signals this faint, MEG uses sensors called SQUIDs (Superconducting Quantum Interference Devices), which must be cooled to -269 degrees Celsius (4 Kelvin) using liquid helium to achieve superconductivity. The entire helmet-shaped MEG scanner sits inside a magnetically shielded room (MSR) that blocks environmental interference. MEG's spatial advantage: unlike electrical signals, magnetic fields are not distorted when they pass through the skull. This means MEG can localize neural sources to within 2-3 mm under ideal conditions — roughly ten times better spatial resolution than EEG. MEG's limitations: equipment costs $2-4 million and requires specialized shielded rooms. The sensors also only detect sources oriented tangentially to the skull, so deep brain structures like the thalamus and hippocampus (which produce mostly radial fields) are largely invisible to MEG.

Researchers choose between EEG and MEG based on their specific scientific question. Temporal resolution: Both are excellent — millisecond precision that fMRI and PET cannot match. Spatial resolution: MEG wins clearly, approximately 2 mm versus 1-2 cm for EEG. Cost and portability: EEG wins dramatically. A research-grade 64-channel EEG system costs around $30,000 and can be used in the field. MEG requires $2-4 million and a fixed shielded room. Deep brain structures: EEG has a slight advantage over MEG for radially oriented deep sources, though neither approaches the depth sensitivity of fMRI. Movement tolerance: EEG can record during walking or driving; MEG requires participants to sit very still inside the fixed sensor helmet. In practice, the most powerful approach combines EEG with fMRI simultaneously — gaining temporal precision from EEG and spatial precision from fMRI. EEG-MEG combinations are also used in clinical epilepsy mapping before surgery.