EEG electrodes must perform two functions. They need to conduct electric signals with a consistent, low impedance connection to the scalp. They also have to stick on and stay in place. The technology that has evolved for this purpose is relatively simple and effective. For routine scalp recordings, cup-shaped electrodes of 1 cm diameter with a flattened rim and a hole in the raised hemisphere are used to ensure good physical contact with the scalp and an enclosed space for electroconductive gel to create a low impedance electric connection. Electrodes are often made of gold, silver, or silver oxidized with chloride (which ensures good ion exchange with the chloride ions in the gel). The technician scrubs and mildly abrades the surface of the scalp at the electrode site with an abrasive detergent paste to remove any dead skin, oil, or dirt. Scalp electrodes should have a maximum impedance of <5000 Ω. EEG technicians check impedance before starting the recording and sometimes during the recording, particularly if an electrode appears to be “misbehaving.” High impedances result in increased noise in channels connected to that electrode. Other problems associated with electrodes can occur if the patient sweats excessively or if the technician inadvertently allows conductive gel to spread between electrodes. Both of these conditions lead to “salt bridges,” low impedance connections between electrodes that transmit low frequency, drifting baseline artifact.

The electrode is attached with either electroconductive paste or collodion, a mixture of nitrocellulose, diethyl ether, and ethanol, which rapidly dries to a stiff, longer-lasting attachment for prolonged recordings. Sometimes, a strip of gauze impregnated with paste or collodion is placed over the electrode and dried onto it by airflow or suction for a very secure connection. Prepositioned electrodes in rubber or elastic caps that stretch over the head should usually be avoided due to the imprecise positioning of the electrodes and often poor electric connections.

Collodion has its proponents and detractors. It is irritating to the skin and mucous membranes, flammable or even explosive, and possibly teratogenic and thus should not be used on a patient with a delicate skin or who might be pregnant. Good ventilation is necessary to avoid inhalation of the ether/alcohol fumes. Collodion does provide a very secure connection; in fact, it requires acetone (“ fingernail polish remover”) to dissolve it and remove the electrodes. Other substances have been used for long-term electrode attachment including cyanoacrylate glues (“superglue”) with reasonable success.

To ensure that EEG recordings are reproducible from one laboratory to another (and consistent from one recording to the next), a standard for electrode placement was developed in the 1950s by Dr. Herbert Jasper at the Montreal Neurological Institute.¹ This widely accepted scheme is known as the International 10-20 System of Electrode Placement.² The underlying principle is that accurate measurements of the head using specific identifiable landmarks can be subdivided into smaller distances based on 10% or 20% increments of the total distances. The three primary measurements are in the sagittal, coronal, and horizontal planes (see Fig. 3.1). The sagittal measurement is from the nasion (notch at the top of the nose) to the inion (midline prominence at the occipital pole). The coronal measurement is from just anterior to the tragus of the ear (in front of the auditory canal) to the midpoint of the sagittal measurement to the opposite tragus. The intersection of these two lines is defined as the vertex. The horizontal baseline is made by connecting points that
are 10% up from the nasion, inion, and tragus points on the sagittal and coronal lines. This circumference measurement defines the horizontal plane.

The electrode positions are determined from 10% to 20% divisions of these measurements. By convention, electrodes are named using capital letters to designate the underlying brain lobe or region (F, frontal; C, central; P, parietal; O, occipital; T, temporal) followed by numbers or lowercase letters to indicate the position in that region. Odd numbers indicate the left hemisphere, even numbers indicate the right hemisphere, and the letter “z” (for “zero”) denotes midline electrodes. “Fp” stands for frontopolar (prefrontal). The numbers increase as the position is more lateral or posterior. The positions of the standard electrodes are given in Table 3.1.

Additional positions can be defined by further dividing these positions into 10% subdivisions (halfway between the 10 and 20 positions), which is occasionally helpful for localization of spike or seizure discharges,³ but the electrode positions described above are usually sufficient for diagnostic purposes and provide excellent coverage for most applications. The standard arrangement for the 10-10 positions is shown in Figure 3.1C. The nomenclature changes slightly for some electrodes.^* The positions are easy to determine and extremely reproducible, requiring only a tape measure and a red wax pencil for marking positions.

The last two lines of Table 3.1 are electrodes that do not fall into the 10-20 system rules but are frequently used to sample brain regions poorly seen by standard electrodes, the anterior and mesial temporal lobes. Sphenoidal electrodes are thin wires introduced by a needle into the region below the zygomatic arch and just above the mandibular notch, about 3 cm anterior to the tragus of the ear. They penetrate about 3 cm deep under the sphenoid wing where they are close to the anterior and mesial temporal lobe.

Whether sphenoidal electrodes actually provide an advantage over external T1/T2 electrodes is a subject of continued debate and research. In a series of 122 patients undergoing long-term monitoring prior to epilepsy surgery, about one-third of patients had EEG findings from sphenoidal electrodes that were not evident in conventional scalp electrodes, and 25% of those with bilateral mesial temporal sclerosis could be selected for surgery on that basis, avoiding the need for intracranial monitoring.⁴ Sphenoidal electrodes also changed the results of source localization of spikes and seizure onsets relative to scalp-only recordings.⁵ By contrast, others have argued that anterior temporal electrodes detect interictal and ictal epileptiform phenomena almost as well as sphenoidal electrodes, provide consistent recordings, do not require physician expertise for placement, and create no discomfort.⁶ Many labs continue to use sphenoidal wires in long-term epilepsy monitoring for localizing seizure onsets.

Another strategy for mesial temporal recording involves percutaneous placement of electrodes through the foramen ovale under fluoroscopic guidance, which picks up some interictal mesial temporal discharges not seen on scalp or sphenoidal recordings.⁷ Nasopharyngeal electrodes placed through the nares into the posterior upper pharynx may also provide a window on the mesial temporal lobe⁸ but are uncomfortable and seldom used.

It is also routine in many laboratories to place “stick-on” electrodes at the right upper and left lower lateral canthus of the eyes, to track electric potentials generated by eye movements (which we will discuss in some detail below). Another pair of electrodes is usually placed on the chest wall and shoulder to record the EKG potentials. Monitoring the EKG is important, as the QRS complex is sometimes seen as an artifact in EEG recordings (particularly in montages with longer interelectrode distances (such as referential montages; see below) and at high sensitivity settings; thus, it helps to have a trace of the EKG activity for comparison. A variety of additional electrodes and transducer inputs (respiratory airflow, EMG, chest movement, etc.) can also be used for evoked potentials, polysomnography, and other specific recording situations.