Electroencephalography (EEG)

CHAPTER 18

I. Physics and Biology of Electricity

A. Ion fluxes and membrane potentials

1. Most of the charge movement in biologic tissue is attributed to passive properties of the membrane or changes in ion conductance

2. Positive ions/cations: K+, Na+, Ca²⁺

3. Negative ions/anions: Cl^–, protein moieties

4. Resting membrane potential: –70 mV (due to difference in permeability of ions and sodium–potassium pump forcing K⁺ in and Na⁺ out)

B. Action potential (AP): an AP normally develops if the depolarization reaches the threshold determined by the voltage-dependent properties of the sodium channels. An AP is based on sodium inward currents and potassium outward currents through voltage-dependent channels.

C. Synaptic transmission

1. Intraneuronal negative polarity of 70 mV noted with intracellular recording.

2. Resting membrane potential is based on outward K⁺ current through passive leakage channels.

3. If resting membrane potential is diminished and threshold is surpassed, the AP is generated.

4. The AP is conducted along the axons to the terminations.

5. When AP reaches presynaptic region, it causes release of neurotransmitter (NT).

6. NT binds to positive-synaptic receptors, opening positive-synaptic membrane channels.

7. Depending on the ionic currents flowing through the transmitter (ligand)-operated channels, two types of positive-synaptic potentials are generated.

a. Excitatory positive-synaptic potential (EPSP)

i. Occurs when sodium inward current prevails

ii. Increases the probability that AP will be propagated

b. Inhibitory positive-synaptic potential (IPSP)

i. Occurs when potassium outward current or chloride inward current prevails

ii. Causes hyperpolarization of the positive-synaptic membrane, making it more difficult to reach the threshold potential

8. Summation

a. EPSPs and IPSPs interact to determine whether AP is propagated positive-synaptically.

b. Temporal summation: EPSPs/IPSPs sequentially summate at a monosynaptic site.

c. Spatial summation: EPSPs/IPSPs simultaneously evoke an end-plate potential polysynaptically.

9. Depolarization of the nerve terminal results in opening of all ionic channels, including those for calcium; calcium entry causes release of NT from the presynaptic terminal that binds to positive-synaptic receptor sites.

10. Chemical transmission is the main mode of neuronal communication and can be excitatory or inhibitory (if positive-synaptic binding opens sodium channels and/or calcium channels → EPSP; if opens potassium channels and/or chloride channels → IPSP); most common excitatory NT is glutamate, common inhibitory NTs are γ-aminobutyric acid (GABA) and glycine.

11. Several EPSPs may be necessary to generate depolarization.

12. Summation of EPSPs in the cortex occurs mainly at the vertically oriented large pyramidal cells.

13. EEG waveforms are generated by the summation of EPSPs and IPSPs that are synchronized by the complex interaction of large populations of cortical cells (but, rhythmic cortical activity is believed to arise from subcortical pacemakers, including the thalamus).

D. Glial cells:

1. Do not generate AP or postsynaptic potentials

2. Resting membrane potential is based only on potassium outward current through leakage channels.

3. With an increase and a subsequent decrease in extracellular potassium concentration, glial cells depolarize and repolarize, respectively.

E. Field potentials and volume conduction

1. Excitatory synapse: the resulting net influx of cations leads to depolarization of the membrane, leading to an EPSP. An intracellular electrode notes the interior becoming more positive than it was at rest, whereas an extracellular electrode sees this as a negative potential.

2. Inhibitory synapse: there is an outflow of cations or an inflow of anions at the synaptic site. Then, the membrane potential is increased at the synaptic site, and a potential gradient develops along the cell membrane EPSPs. The electrode near the synapse “sees” a positivity and the electrode distant from the synapse a negativity.

3. Field potentials are generated by extracellular currents. Negative field potentials at the cortical surface may be based on superficial EPSPs as well as on deep IPSPs, and positive field potentials at the surface may be based on superficial IPSPs as well as on deep EPSPs.

4. The movement of charge from excitable tissue to surrounding tissue is called volume conduction.

F. Generation of EEG rhythms

1. Cortical potentials

a. Electrical activity in the deep cortical nuclei produces surface potentials of low amplitude.

b. The largest neurons are involved in efferent outflow and are oriented perpendicular to the cortical surface, producing a vertical columnar orientation of the cortex.

c. Influx of positive ions into the efferent neurons results in a negative extracellular field potential; electrotonic depolarization of the soma and axon hillock results in a positive field potential; because of the vertical orientation of the large efferent neurons, the negative field potential is usually superficial to the positive field potential, forming a dipole.

d. In humans, the thalamus is believed to be the main site of origin of EEG rhythms; oscillations at the thalamic level activate cortical neurons; EPSPs acting on the dendrites mainly in layer 4 (the main site of depolarization) create a dipole with negativity at layer four and positivity at more superficial layers; scalp electrodes detect a small but perceptible far-field potential that represents the summed potential fluctuations.

2. Scalp potentials

a. It is estimated that 6 to 10 cm² of cortex must be synchronously activated for a potential to be recorded at the scalp (note: potentials must be volume conducted through the meninges, skull, and skin before being detected by scalp electrodes).

b. The source dipole is perpendicular to the surface. Scalp potentials are determined by the vectors of cortical activity; if the superficial layer four of the cortex is a positive field potential and deeper layers are negative, then there is a vertical vector produced, with the positive end pointing toward the scalp electrode; the amplitude of the vector depends on the total area of activated cortex and the degree of synchrony among the neurons.

c. The source of the electrode is near to the surface. Scalp electrodes can record a few millimeters deep and are not able to detect deep nuclei; scalp EEG, therefore, records approximately one-third of cortical activity.

d. The head is a uniform and homogeneous volume conductor.

3. At least one recording electrode is essentially over the source and the reference is not contained in the active region.

G. Generation of epileptiform activity

1. Generated when depolarization results in synchronous activation of many neurons

2. Spikes and sharp waves

a. Duration: spikes, <70 milliseconds

b. Duration: sharp wave, 70 to 200 milliseconds

c. Spike potentials are the summation of synchronous EPSPs and APs in the cortex; the foundation for the bursting spike potential is the paroxysmal depolarization shift.

d. The negative end of the epileptiform dipole points toward the cortical surface, resulting in a negative deflection at the scalp electrode.

e. The distribution of the epileptiform potential across the cortical surface is called the field.

f. Occasionally, the surface is positive (and in normal patterns of positives and 14- and 6-Hz–positive spikes).

3. Paroxysmal depolarization shifts

a. Extracellular field potentials characterized by waves of depolarization followed by repolarization

b. High-amplitude afferent input to the cortex produces depolarization of cortical neurons sufficient to trigger repetitive APs, which in turn contribute to the potentials recorded at the scalp rhythmicity, likely due to a mechanism inherent of neurons to be unable to sustain prolonged high-frequency discharges (termination of the sustained depolarization is likely due to activation of K⁺ channels and inactivation of Ca²⁺ channels).

c. Ultimately, termination of epileptiform discharges is due to inhibitory feedback to neurons.

d. Note: the previous points are for partial seizures ± secondary generalization; for primary generalized seizures, the generator is likely a loop between the cortex and thalamus (possibly also responsible for sleep spindles).

II. The EEG Machine, Electrodes, and Their Derivations

A. Input board: channel—formed by the two selected electrodes, amplifier, and recording unit to form a system to display the potential differences between two electrodes

B. Filters

1. Filters selectively reduce the amplitude of voltage changes or signals of selected frequencies

2. Types of EEG filters

a. Low-frequency filter (LFF; also known as high-pass filter)

i. Allows frequencies higher than designated

ii. Typically maintained between 0.5 and 1.0 Hz

b. High-frequency filter (HFF; also known as low-pass filter)

i. Should not have HFF <30 Hz on scalp EEG because high-frequency epileptiform discharges may be filtered.

c. 60-Hz filter (to attenuate artifact caused by electrical power lines)

C. Amplifiers

1. Amplifier sensitivity is typically 7 µV/mm.

2. Two main functions of the amplifier: discrimination and amplification

3. Each amplifier has two inputs connected to the input selector switches.

4. EEG amplifiers are differential amplifiers (increase the difference in voltage between the two input terminals, with identical inputs of the two terminals being rejected); this serves to distinguish cerebral potentials that are likely to have different amplitude, shape, and timing at electrodes in different regions and allows rejection of potentials that will be similar at all electrodes (e.g., 60 Hz) if impedance is equal at all electrodes; failure to reject artifact, such as 60 Hz, may occur if impedance is different at the two input electrodes or there is absence of an effective ground to the patient.

D. Calibration

1. Square-wave calibration: square-wave pulse of 50-µV amplitude is delivered to the inputs of each amplifier at rate of 1-second intervals from square-wave pulse.

2. Biocalibration: assesses the response of the amplifiers, filters, and so forth, to complex biologic signals

E. “Paper” speed: typically 30 mm per second

1. Electrodes

2. Usually made of gold, silver chloride, or other material that does not interact chemically with the scalp; skin is prepared by abrasion to remove excess oils and dead skin containing low levels of electrolytes that may alter impedance; electrode gel (usually NaCl) is used to reduce resistance and improve contact of the electrode to the skin.

3. Impedance between the scalp and electrode must be <1,000 Ω.

4. Electrode placement

a. Standard 10 to 20 international system

i. Measuring the head

(A) Measure from nasion to inion and mark at 50% point.

(B) Measure between the two preauricular points and mark at 50%; the intersection with step 1 is CZ.

b. Intracranial electrodes

i. Depth electrodes

ii. Subdural/epidural grids and strips

III. Montage

A. Referential

1. Localizes epileptiform potentials by amplitude and complexity (sharpness) of the waveform.

2. Interelectrode distance alters amplitude; usually CZ or CPz (midline posterior electrodes) are used as reference electrodes or ipsilateral (IL)/contralateral (CL) ear electrodes.

3. Digital EEG allows for average of electrodes as reference.

B. Bipolar: phase reversal—localization based on positive deflection in one channel with negative deflection in adjacent channel (epileptiform potentials on scalp recordings are typically electronegative)

IV. Rhythm and Frequency

A. Frequency categories

1. ∆: 1.0 to 3.9 Hz

2. θ: 4.0 to 7.9 Hz

3. α: 8.0 to 12.9 Hz

4. β: ≥13 Hz

B. Normal awake rhythm in adult

1. Rhythm and frequency differ in that rhythm is a subcortical generation (likely thalamus) of continuous activity, whereas frequency describes that rate at a given time for recorded activity.

2. α rhythm with attenuation/reactivity with eye opening

V. Artifact

A. Physiologic artifacts: usually due to movement, bioelectric potentials, or skin resistance changes

1. Eye movement: cornea approximately 100-mV positive compared to retina

2. Cardiovascular: often noted in temporal electrodes

3. Perspiration: causes slow waves usually >2 seconds in duration (0.5 Hz) owing to changes in impedance between electrode and skin

4. Muscle: usually ≥35 Hz

5. Galvanic skin response: slow waves of 1 to 2 Hz that last for 1 to 2 seconds with two to three prominent phases; represents an autonomic response of sweat glands and changes in skin conductance in response to sensory stimulus or psychic event

B. Nonphysiologic artifacts

1. Two main sources

a. External electrical signal: 60-Hz electrical input; factors that reduce 60-Hz artifact:

i. Proper ground

ii. Keeping electrode impedance low and approximately equal

iii. Keeping power lines away from electrodes

iv. Shielded room to reduce artifact from electricity

b. EEG equipment: electrode pops

i. Spike-like potentials that occur in random fashion and are caused by sudden changes in junction potentials

ii. Small movements or alterations of the electrode–gel interface may temporarily short out the junction potential, and the sudden change in junction potential is seen in all channels with that electrode in common.

iii. Dissimilar metals build up large junction potentials that are discharged into the amplifier.

C. Procedures

1. Hyperventilation (HV)

a. Duration: at least 3 minutes of adequate effort (5 minutes if absence seizure is suspected)

b. Useful in primary generalized seizure disorders; HV will elicit seizure activity in 75% of patients with absence seizures.

c. HV not performed in elderly or other patients with possible cardiovascular/atherosclerotic disease owing to the risk of vasoconstriction with resultant cardiac or cerebral hypoperfusion.

2. Photic stimulation

a. Useful in primary generalized seizure disorders

b. May demonstrate occipital driving (typically at photic frequency near baseline background cortical frequencies)

3. Sleep deprivation

a. Potential for increased epileptiform activity in light sleep

b. Useful in primary generalized seizure disorders and partial seizure disorders

VI. Normal EEG Findings

A. Normal background cortical activity

1. During the awake state, the patient demonstrates a well-modulated, well-developed 8- to 10-Hz posterior predominant α rhythm.

2. Attenuates with eye opening

B. Sleep patterns: divided into non-rapid eye movement (NREM) and rapid eye movement (REM)

1. NREM sleep:

a. Stage I:

i. Drowsiness

ii. About 5% of sleep time is spent in stage I sleep.

iii. Slow eye movements (less than 0.5 Hz)

iv. Attenuation of background and alpha rhythms replaced by theta activity

v. Enhancement of beta activity in the fronto/central regions.

b. Stage II:

i. Light sleep

ii. About 45% of time is spent in stage II sleep.

iii. Characterized by sleep spindles (11- to 16-Hz, localized in the fronto-central region) and/or K complexes (vertex waves followed by spindles)

iv. Positive occipital sharp transients of sleep (POSTS): triangular waves at irregular intervals (usually >1 second)

v. Vertex (V) waves: negative sharp transients at vertex

c. Stage III (previously divided into III and IV):

i. Deep sleep

ii. 25% of total sleep

iii. POSTS, sleep spindles, K complexes

iv. Delta activity

2. REM sleep

a. Low-amplitude EEG

b. Sawtooth waves (medium amplitude, theta waves) over central region

c. Decreased muscle activity

C. µ Rhythm