EEG · Grey Matter

EEG (electroencephalography) records the brain's electrical activity through electrodes on the scalp, and it is the cheapest, most portable way to follow the brain in real time, used to diagnose epilepsy, stage sleep and drive simple brain-computer interfaces.

EEG. EEG (electroencephalography) records the brain's electrical activity through electrodes on the scalp, and it is the cheapest, most portable way to follow the brain in real time, used to diagnose epilepsy, stage sleep and drive simple brain-computer interfaces.

What it measures is narrower than the name suggests. The signal comes almost entirely from the slow synaptic currents (excitatory and inhibitory postsynaptic potentials) of pyramidal cells in the cortex. These cells stand side by side with their dendrites pointing to the surface, so when thousands of them receive input together their currents add up into a dipole large enough to be seen through the skull; the brief action potentials are too fast and too poorly aligned to add up. A deflection on the scalp needs something like 6 to 10 square centimetres of cortex active in synchrony.

EEG sees synchrony near the surface. Deep structures, asynchronous activity and sources whose currents cancel are nearly invisible, so a flat stretch of EEG does not mean a silent brain.

The skull blurs it. Bone conducts poorly, so the potential spreads sideways before it reaches the scalp, and each electrode sums a wide patch of cortex; locating a source from the electrodes is an inverse problem with no unique answer.

Its strength is time. EEG follows changes within milliseconds, sampled at hundreds of hertz or more, while blood-flow methods such as fMRI follow them over seconds.

The signal is microvolts, and the eyes, muscles and mains wiring produce larger ones; much of EEG practice is about recognising and removing these artefacts.

Listening to a stadium from the car park is the picture of EEG.

A chant of thousands comes through clearly, a single conversation never does, and the walls muffle and spread everything.

Questions: What does a seizure look like on an EEG, and what can the EEG miss? During a seizure the trace changes from low, irregular activity to rhythmic discharges that grow in amplitude, change frequency over seconds and spread to neighbouring electrodes, followed after the seizure by slowing. Between seizures, brief sharp waves and spikes in one region can mark tissue prone to seizures. Because the scalp EEG needs several square centimetres of synchronous cortex, a small or deep seizure can be invisible on it, which is one reason electrodes are sometimes placed inside the skull to find where seizures start. How small is the brain's signal at the scalp compared with the noise around it? Scalp EEG is measured in tens of microvolts, and it already reflects several square centimetres of cortex firing in step; a single neuron contributes nothing measurable at that distance. Blinks, eye movements, jaw and neck muscles and the heart produce signals of similar or larger size, and mains wiring adds interference at 50 or 60 Hz, so much of EEG practice is about recognising and removing those artefacts. The noise does not shrink as fast as the brain's signal does with distance, which is why recording closer to the source improves signal more than any better amplifier. What does it mean when an EEG shows strong alpha or strong delta? An EEG is usually read by splitting it into bands: delta below about 4 Hz, theta about 4 to 7 Hz, alpha about 8 to 12 Hz, beta about 13 to 30 Hz and gamma above about 30 Hz, with edges that vary slightly between textbooks. Strong alpha over the back of the head is the normal pattern of an awake person at rest with the eyes closed, and it shrinks when the eyes open. Strong delta is normal in deep sleep and abnormal in an awake adult, where it can point to a damaged or depressed patch of cortex. A band's power says how many neurons are oscillating together at that pace, so it describes the state of the tissue rather than any specific thought. What does a spike on the EEG between seizures say about the brain? An interictal spike, a sharp wave of 20 to 70 ms standing out from the background, shows that a patch of cortex can make many neurons fire almost at once, the same synchrony a seizure is built from, while inhibition still cuts it short. It supports a diagnosis of epilepsy and its position on the scalp points to where focal seizures may begin, though it causes no symptom by itself. Its absence on a single recording proves little, and long recordings from implanted devices show spike rates rising and falling over cycles of weeks, with seizures more likely near the peaks. What does moving the electrodes from the scalp to the surface of the brain gain? It removes the skull, the layer that most attenuates and blurs the signal. Electrodes on the cortex record from a patch millimetres across instead of centimetres, see amplitudes several times larger, pick up far fewer muscle and eye artefacts, and keep the high-frequency activity (above about 70 Hz) that tracks local firing and that the skull filters out. The price is surgery and limited coverage, since a grid only sees the cortex it lies on, and it still records populations of neurons, never single cells. What does MEG see that EEG does not, and the other way round? Both record the same synchronous currents of pyramidal cells, but the skull treats them differently: it smears the electric potential that EEG measures and leaves the magnetic field almost untouched, so MEG maps sources more sharply. MEG is mostly blind to currents pointing straight out of the head (the crowns of the cortical folds) and sees best those running parallel to the scalp (the walls of the folds), while EEG picks up both orientations. EEG is cheap and portable; MEG needs a shielded room and, until optically pumped sensors arrived, a helium-cooled helmet. Recording both together gives two views of the same sources and constrains the inverse problem better than either alone. Why does EEG record synaptic currents of pyramidal cells and miss their action potentials? To reach the scalp, currents from thousands of cells have to add up. Pyramidal cells stand in parallel with their dendrites toward the surface, and their synaptic currents last tens of milliseconds, so inputs that arrive roughly together overlap in time and sum into one large dipole. An action potential lasts about a millisecond and its currents flow in opposite directions over a short stretch of axon, so spikes from many cells rarely overlap and largely cancel at a distance. Even so, a visible deflection needs something like 6 to 10 square centimetres of cortex active in synchrony.