ELECTROENCEPHALOGRAPHY

In 1929, Hans Berger, a German psychiatrist, recorded the first human electroencephalogram (EEG). In the 1940s, Gibbs and colleagues demonstrated 3 Hz spike–wave discharge in ‘petit mal’ epilepsy (now ‘absence’ epilepsy). The EEG is crucial for diagnosing, classifying and managing people with seizures and other neurological disorders. Currently, the majority of EEG reporting is based on visual reading by skilled clinical neurophysiologists. AI‐driven EEG reporting is becoming available and is expected to have a significant impact on EEG reporting.

BASIC PRINCIPLES

Electrical activity recorded by scalp electrodes reflects the summation of excitatory and inhibitory postsynaptic potentials in apical dendrites of pyramidal neurons in superficial cortex layers (Figure 7.1a). Most of the signal is produced by vertically oriented neurons.¹ The EEG signal is measured as the potential difference between two electrodes. It is attenuated, filtered, and potentially contaminated by structures between the cortex and the recording electrode, including CSF, dura, bone, muscle, blood vessels, and skin. Six to ten cm² of cortex needs to be activated synchronously to generate enough potential for changes to be recorded from the scalp.²

The electrical signal needs to be amplified by a factor of 100,000, and the recording is prone to physiological artefacts (such as ECG, eye blinks, chewing, or other muscle activity) and non‐physiological artefacts (electrical interference). Scalp EEG has high temporal resolution but low spatial resolution, as significant amounts of the cortex, particularly in basal and mesial areas of the hemispheres, are not covered by standard electrode placements.

A standard routine EEG recording takes about an hour. Electrodes are applied using the international 10–20 system, which uses electrodes 10% or 20% of the distance between standard landmarks (Figure 7.1b). Additional electrodes may be placed, depending on specific questions, such as surface electromyographic (EMG) electrodes that enable the correlation of muscle activity with EEG activity. ECG electrodes and respiratory belts are used to examine autonomic function and to detect cardiac arrhythmias, which are a crucial part of differential diagnosis in individuals presenting with blackouts (Figure 7.2). The EEG can be displayed using different montages, such as bipolar, in which each waveform in the EEG represents the difference in voltage between two adjacent electrodes and a reference showing the voltage between a specific active electrode and a designated reference (Figure 7.3).

EEG FREQUENCY

The standard scalp EEG picks up activity between 1 to 30 Hz (delta 1‐4 Hz, theta 4‐7 Hz, alpha 8–12 Hz, beta 13‐30 Hz) (Table 7.1). Faster frequencies known as high frequency oscillations (HFOs) (gamma 30‐80 Hz, ripples 80‐ 250 Hz, and fast ripples 250‐500 Hz) can be picked up with intracranial EEG, which is not attenuated by scalp and skull. These may be a biomarker for epileptogenic tissues, though physiological HFOs have also been noted in people without epilepsy. Higher‐density scalp EEG, when combined with computational analysis, may provide additional information on the source localisation of activity in focal epilepsy.

ACTIVATION TECHNIQUES

Hyperventilation and photic stimulation increase the yield of abnormal findings and are included in routine EEG following standardised protocols. These procedures are generally safe; however, there are specific contraindications to consider. Hyperventilation and photic stimulation are associated with a small risk of triggering seizures.

The National Institute for Health and Care Excellence (NICE) guidelines recommend hyperventilation for a minimum of 3 minutes, if not contraindicated. Hyperventilation may trigger a seizure in a small number of people, so the individual should provide informed consent before undergoing this procedure.³ It may not be advisable to use hyperventilation in seizure‐free individuals. It should also be avoided in those with significant pulmonary and cardiac disease or raised intracranial pressure. It can be instrumental in the diagnosis of absence epilepsies, as well as for triggering non‐epileptic events.

During intermittent photic stimulation, the individual is presented with a series of trains of flashing lights. The photoparoxysmal response occurs when photic stimulation produces bilaterally synchronous epileptiform discharges that outlast the stimulus. The middle flash frequencies, between 10 and 20 Hz, are most provocative. This is most commonly seen in genetic generalised epilepsy but can be seen in some focal epilepsies, and may also exist as a genetic trait in some people without epilepsy.⁴

EEG IN CLINICAL PRACTICE

ICTAL EEG

Routine EEG recordings are unlikely to capture events, and seizures are usually only captured during longer recordings. The exception is absence seizures, which can be triggered by hyperventilation. Ictal patterns are identified by the abrupt onset of epileptiform activity that evolves in frequency, amplitude, and /or spatial distribution and terminates abruptly. Some seizure types have characteristic patterns, such as the 3 Hz spike and wave in absence seizures, the evolving 5–7 Hz temporal theta rhythm in mesial temporal epilepsies, and high‐frequency discharges in tonic and neocortical seizures. Ictal changes may be obscured by artefacts from movement or muscle, particularly in frontal lobe epilepsies with motor features, and also due to the distance from the recording electrodes on the scalp. Scalp EEG commonly shows no change in focal preserved consciousness seizures (auras), because the ictal discharge involves too small a neuronal aggregate for the synchronised activity to be detected on the scalp.