Basic Theory of EEG

Abstract

EEG is the most sensitive method to reflect the functional state of the brain. In neurocritical care units (NICUs), most patients often cannot express their feelings because of consciousness disorders, clinical physical examination and imaging examination cannot achieve continuous evaluation, and routine electrocardiogram (ECG)-blood pressure-cerebral oxygen saturation monitoring has difficulty reflecting the changes in patients’ brain function sensitively, which easily leads to a delay in treatment and the deterioration of the patient’s condition; therefore, continuous and sensitive (EEG) monitoring is important for neurocritical patients. On the other hand, the EEG waveform is complex because different patterns may represent different stages of a disease, and even different sleep-wake states in the same stage of a disease may also lead to deviation in the EEG interpretation. In the NICU, many monitoring instruments and clinical operations can easily lead to various EEG artifacts. How to explain the significance of changes in EEGs reasonably and distinguish artifact interference accurately will have an important impact on the decisions of clinicians. This chapter briefly introduces the basic concepts of EEG, normal EEG waves, and key points in EEG interpretation and emphasizes the common artifact recognition techniques in the NICU. Throughout the chapter, we put forward our own views, and we hope to provide a solid foundation for the contents described in the following chapters in this book.

EEG is a method to record and trace the bioelectrical activity of brain neurons. Since the German neuropsychiatrist Hans Berger began to study human EEG in 1924, EEG has been developed all over the world and began to serve clinical and scientific services. The International Society of EEG and Clinical Physiology was established in 1947, and the first International EEG Conference was held in the United Kingdom. In China, in the early 1980s, Beijing took the lead in establishing the epilepsy and EEG group, followed by the establishment of corresponding organizations in various provinces and regions across the country. On this basis, the National Society of EEG and Epilepsy was born, and its meeting was held every 2 years. Currently, EEG examination has been popularized in hospitals throughout the world as a mature technology, serving the majority of patients. After the 1970s, with the development of electronic technology, dynamic cassette EEG, EEG video monitoring systems, digital EEG, and stereotactic EEG came into being, providing more opportunities for research in the fields of epilepsy, epileptic seizure-type diagnosis, and sleep physiology and pathology.

1 Formation of EEG

The human brain, similar to other parts of the body, such as the heart and muscles, can generate bioelectricity. The recording of brain bioelectrical activity by placing electrodes on the scalp is called EEG. However, the current generated by the human brain is very weak; therefore, it must be amplified one million times, and through electromagnetic induction, the pulsed direct current traced from the scalp electrodes is converted to alternating current (AC), and then by multipole amplification, recorded on record paper. Therefore, the commonly seen EEG is an indirect image of brain electrical activity.

2 Mechanism of the Formation of an EEG

The human brain includes four parts: cerebrum, diencephalon, cerebellum, and brain stem. The brain is divided into two cerebral hemispheres, which are connected together by the corpus callosum and the anterior and posterior commissures of the brain. The outermost surface of the cerebral hemisphere is gray matter, which is mainly composed of nerve cells, and it is called the cerebral cortex. The deep part of the cerebral hemisphere is composed of fibers of nerve cells called white matter. The cerebral hemisphere is divided into four lobes: the frontal, parietal, occipital, and temporal lobes. The diencephalon is located between the two cerebral hemispheres above the midbrain, including the thalamus, hypothalamus, and third ventricle; and the brainstem is divided into the midbrain, pons, and medulla oblongata.

The cerebral cortex is the highest center of human advanced neural activity and the main anatomical basis of brain-wave activity. The human cerebral cortex is composed of approximately 14 billion neurons and countless synapses that form a large and complex information transmission network. The cerebral cortex is divided into six layers, from the outside to the inside: (a) molecular layer: there are horizontal cells inside, and the axons within this layer run horizontally on the surface of the cortex with lateral conduction function; (b) outer granular layer (small pyramidal cell layer); (c) pyramidal cell layer: the apical dendrites are as long as the cortical surface; (d) inner granular layer (astrocytic layer); (e) ganglion cell layer; (f) fusiform or polymorphic cell layer: the axons of this layer extend to adjacent white matter. The central nervous system (CNS) is basically composed of nerve cells (neurons), glial cells, and nerve fibers. Nerve cells are composed of cell bodies and protrusions extending from the cell bodies. The latter are divided into two types, dendrites and axons, according to their functions. Each neuron has multiple dendrites, and each dendrite branches repeatedly to form a branch shape. Dendrites are the most active part of nerve excitement transmission, and they receive external impulses. Dendrites are divided into short dendrites and apical dendrites; the former connect with neurons, and the latter extend to the surface of the cortex. Each neuron has only one axon, whose function is to transmit nerve impulses from the cell body. Nerve impulses are not directly transmitted from one neuron to the next but must pass through a contact point, called a synapse. The terminus of a neuron is divided into many small branches, and the end of each small branch expands into a spherical shape, called a synaptosome, which is attached to the cell body or dendritic surface of the next neuron. At the contacts of the synapses, each is separated by a membrane. The axon membrane at the end of the axon is called the presynaptic membrane, the cell body membrane or dendritic membrane opposite to it is called the postsynaptic membrane, and the gap between the two membranes is called the synaptic cleft. There are dense protrusions on the inner side of the presynaptic membrane that form a vesicle fence. The synaptic corpuscle contains a large number of synaptic vesicles containing neurotransmitters. Synaptic vesicles release transmitters through the vesicle barrier. In some parts of the postsynaptic membrane, there is a special substance that can bind to neurotransmitters called receptors.

Brain electrical activity originates from the cerebral cortex, and brain electrical activity is generated by the postsynaptic potential (PSP) of pyramidal neurons and their vertical apical dendrites.

In the static state, if a microelectrode is inserted into a cell, there will be a potential difference between the inside and the outside of the membrane. The potential inside the cell membrane is low, 60–90 mV negative relative to the outside of the membrane. If no stimulation is given to neurons, this potential difference remains constant. This potential difference between the inside and the outside of the membrane is called the resting potential or membrane potential. The formation of membrane potential is related to the concentration of various ions in the fluid inside and outside the cell. Under normal circumstances, the main positive ion in the intracellular fluid is K⁺, and the main negative ion is organic acid. In the extracellular fluid, the main positive ion is Na⁺, and the main negative ion is Cl⁻. The cell membrane has certain selectivity for ion permeability. In the static state, the permeability to K⁺ is the largest, followed by Cl⁻; the permeability to Na⁺ is very small, only 1/50 that of K⁺, and the cell membrane is completely impermeable to organic matter. Ions always diffuse from places with high concentrations to places with low concentrations; therefore, K⁺ diffuses out of the cell through the cell membrane, while the organic anions inside the cell stay inside the cell. It is difficult for Na⁺ outside the cell to diffuse into the cell through the cell membrane. Therefore, there are many positive ions outside the cell membrane, and the potential is high; there are many negative ions inside the cell membrane, and the potential is low. This phenomenon of polarity is called the polarization state. Because the positive and negative charges inside and outside the membrane attract each other, the positive and negative ions are arranged outside and inside the cell membrane, so a potential difference is formed inside and outside the membrane, which is the static potential mentioned above.

When a neuron is excited, the permeability of the cell membrane to ions is changed, and the permeability of the cell membrane to Na⁺ is selectively increased. Therefore, Na⁺ diffuses from the outside of the cell to the inside of the cell, increasing the positive ions in the membrane and canceling the original membrane potential, which is called depolarization. Finally, the potential inside the cell is even higher than that outside the cell. The potential change generated at this time is called the action potential (AP). When the AP reaches the peak, the permeability of the membrane to Na⁺ decreases, and the permeability to K⁺ increases significantly, so K⁺ diffuses out of the membrane and returns to a positive polarization state outside the membrane and a negative polarization state inside the membrane, which is called repolarization. In the process of repolarization, with the help of the Na-K pump of the cell membrane, the excess Na⁺ that has diffused into the cell is transported to the outside of the cell, and the extra K⁺ is transported to the cell. Since the solutions on both sides of the cell membrane are conductive, a circular current loop is formed between the nerve segments in the excitatory and resting parts. This circular current loop depolarizes the previously dormant nerve membrane in the adjacent region, forming a new excitation region. A local circuit is formed between the new excitement area and the next adjacent part. Repeatedly, the excitement spreads along the nerve fiber.

Nerve impulses are not directly transmitted from one neuron to the next but are connected in the form of synapses. When the nerve impulse is transmitted from the presynaptic neuron to the postsynaptic neuron to the synapse, transfer mediators (such as acetylcholine and γ-aminobutyric acid (GABA) stored in the synaptic corpus are released and act on the postsynaptic membrane through the synaptic cleft. The transmitter binds to receptors in the postsynaptic membrane, which temporarily changes the permeability of the postsynaptic membrane to ions, changing the membrane potential and generating a local current. When the local current reaches certain intensity, the excitement is transmitted to the next neuron, and this potential is called the PSP. If the chemical transmitter released from the synaptic corpuscle is acetylcholine, it will increase the permeability of the postsynaptic membrane to Na⁺, K⁺, and Cl⁻, but the permeability to Na⁺ is the greatest, causing a depolarizing PSP, namely, an excitatory postsynaptic potential (EPSP); an EPSP is a temporary partial reduction of membrane potential. If synaptic vesicles release GABA, the permeability of the postsynaptic membrane to K⁺ increases, causing excessive polarization of the postsynaptic membrane, which is called an inhibitory postsynaptic potential (IPSP). An IPSP is a temporary increase in membrane potential. The potentials generated by the synapses in different parts of the cell are summed up in the membrane potential of the cell body.

The total potential of the cerebral cortex mainly occurs in the large pyramidal cells arranged in the vertical direction in the cortex. Most of the current is limited to the cortex, and a small part of it passes through the meninges, cerebrospinal fluid (CSF), and the head to the scalp, causing different potential levels in different parts of the scalp. The amplitude of these potential differences is 10–100 μV, which can be recorded between two electrodes. This is the basis of EEG [1].

3 EEG Distribution, Waveform, and Amplitude

Since EEG waves represent the synchronous potential difference of a group of nerve cells in a certain area of the cerebral cortex, EEG analysis should be conducted from seven aspects, namely, frequency, amplitude, waveform, phase relation, the mode of occurrence of abnormal waves, the breadth of distribution, and responsiveness to various stimuli.

3.1 Frequency

Frequency refers to the number of times a certain wave repeats in 1 s. It is usually expressed in cycles per second (c/s) or Hz. The wavelength of the scattered slow wave can be measured and expressed by the time it occupies. The frequency of brain waves is divided into four frequency bands: δ frequency band: 3/s or less; θ frequency band: 4–7/s; α frequency band: 8–13/s; and β frequency band: 14/s or more (usually 14–40/s). The frequency band of EEG is one of the important criteria for evaluating abnormalities in clinical practice. Waves below 8/s are called slow waves, and waves above 13/s are called fast waves.

3.2 Amplitude

Amplitude represents the height of a wave, expressed in μV. It is determined by comparing the distance of a wave from its peak to the intersection of the two-point line of the wave and the height of the standard signal recorded under the same gain and filtering conditions. For example, if the height of a wave in an EEG is 10 mm, and the calibration signal of 50 μV is measured to be 5 mm, then the amplitude of this wave is 100 μV. The amplitudes of the corresponding parts on the two sides should be symmetrical. The asymmetry of the amplitudes is often caused by extracerebral factors. In particular, the difference in the amplitude of the recording electrodes can be caused by different spatial distances or impedances. Therefore, the placement and resistance of the electrodes should be verified during the inspection.

3.3 Waveform

• Sinusoidal wave: The ascending and descending branches of a sinusoidal wave are clear and smooth.

• Monophasic and diphasic waves: A monophasic wave is a single-direction deflection upward or downward from the baseline, while a diphasic wave includes two components above and below the baseline.

• Spike wave: A spike wave looks like a spike, and a spike waves takes 20–70 ms.

• Sharp wave: A sharp wave is sharp-like and differs from a spike wave in that it takes more than 70 ms (between 70 and 200 ms), the upward branch is usually steeper, and the downward branch is sloping.

• Complex wave: A complex wave is composed of two or more continuous waves. For example, a spike-slow complex wave consists of a spike wave and a slow wave, such as a 3/s spike-slow complex wave and a 4~5/s spike-slow complex wave. Different complex waves have different clinical meanings.

3.4 Phase Relation

Phase relation refers to the synchronization and polarity relation of brain waves in one or more leads. In different leads, if the peaks and troughs of waves occur at the same time, these waves are said to be in phase; if their peaks and troughs do not occur at the same time, they are said to be out of phase. If the two wave crests are in completely opposite directions and are 180 degrees out of phase, these waves are said to be phase reversed. Phase reversed waves are a sign of the origin of brain-wave potential during bipolar tracing, and they have positioning significance.

3.5 Mode of Occurrence of Abnormal Waves

Long-range: appear continuously for 5–10 s. Short-range: appear continuously for 1–2 s. Rhythmicity: reoccur regularly. Irregularity: highly rhythmic disorder; irregular waveforms, amplitudes, and frequencies; scattered or appear occasionally. Migratory: a wave that occurs at a certain time in a certain area but occurs in another area at another time.

3.6 Distribution and Breadth

• Universal or diffuse distribution Electrical activity that occurs in various areas or most areas of the brain at the same time. It may be that the amplitude is higher in some regions of its distribution.

• Lateral distribution Appears on one side of the cerebral hemisphere or mainly on one side. This distribution is abnormal. The abnormality of the brain is located on the side of the abnormal activity or with a lack of normal activity.

• Focal distribution Electrical activity occurs only in a certain area of the head. The electrodes in the neighboring area may be affected and have the same electrical activity, but the amplitude of the electrical activity in the affected area is relatively low. The focal distribution should be distinguished from the predominance of a certain area accompanied by the universal distribution. For abnormal slow waves and sharp waves, this distinction is very important.

3.7 Reactivity

The normal and abnormal EEG changes induced by various methods are called reactivity. These methods include open and closed eye tests, hyperventilation, flashes or other sensory stimuli, and changes in alertness levels.

4 Common EEG Recording Methods

EEG records and traces the bioelectrical activity of brain neuron cells. The brain electrical activity recorded by scalp electrodes is usually 1–60 Hz, and the voltage is between 5 and 300/μV. The conductors placed on the scalp to export brain electrical activity are called electrodes. There are many types of electrodes. Commonly used scalp electrodes include needle electrodes, tubular electrodes, and disc electrodes; and special electrodes include sphenoid electrodes, nasopharyngeal electrodes, and intracranial electrodes.

For conventional electrode placement, the International Federation of Societies for electroencephalography and Clinical Neurophysiology (IFSECN) recommended the International 10–20 system. Electrodes are generally placed on the left and right forehead (Fp), frontal area (F), central area (C) (anteroposterior to the central groove), parietal area (P), occipital area (O), anterior temporal area (T or F7, 8), middle temporal area (T or T3, 4), posterior temporal area (T or T5, 6), earlobes on both sides (A1, A2), and the middle of the head (parietal center C). Odd-numbered electrodes are placed on the left side of the head, and even-numbered electrodes are placed on the right side of the head. The electrodes are passed through an electroencephalograph composed of several amplifiers to amplify the weak electrical activity of the brain via wires and record this activity by a recording device on paper or a computer screen. The EEG examination room should not be too bright or soundproof, but it should be quiet. There should be glass windows on the shielding wall of the EEG device so that the examiner can see the patient and observe their condition. For example, if a patient has seizures or other conditions during the tracing process, the examiner can see this and give timely treatment.

At least two electrodes are required for EEG. One electrode is connected to the first grid (G1) of the EEG machine and the other is connected to the second grid (G2). The potential difference recorded between the two electrodes forms an EEG. Each amplifier has two input terminals, which are connected to the G1 and G2 to record the potential difference. According to custom, when the potential of G1 is more negative than G2, the recorded waveform is required to be upward (negative phase, negative); conversely, if G1 is more positive than G2, the waveform is downward (positive phase, positive). Therefore, G1 is the negative terminal of the amplifier, and G2 is the positive terminal.

If there is a point with a zero potential on the body connected in parallel to the G2, the potential difference with the electrodes on other parts of the G1 is equal to the absolute value of the latter’s potential change. This zero-potential point theoretically refers to the point where the body is infinitely far away from the body when it is in the electrolyte solution. In fact, this kind of absolute-zero potential does not exist. The lead method (montage) of EEG can be divided into the monopolar lead method (monopolar), using irrelevant electrodes, and the bipolar lead method (bipolar), using only active electrodes without using irrelevant electrodes.

4.1 Monopolar Lead (Fig. 1.1)

The monopolar lead method involves connecting each active electrode of the scalp with an irrelevant electrode on the same side, and the EEG traced is the potential difference between each active electrode and the irrelevant electrode. The irrelevant electrode that is often used is the ear electrode, and the zero potential of the ear electrode is set to indicate the absolute value of the potential of each active electrode on the scalp. However, in fact, the ear poles do not have absolute-zero potential and may be affected by bioelectricity other than brain electricity, such as ECG and myoelectricity. Therefore, the standard monopolar lead traces only the potential difference between the active electrodes of the scalp and the ear electrodes, which is very close in value to the absolute value of the active electrode potential. It is recommended to use average monopolar leads at the same time; that is, irrelevant electrodes use the average value of each scalp electrode potential through high-resistance input (0.5~3 M) as a reference to replace the ear electrode to eliminate the influence of the ear electrode.

Fig. 1.1

Monopolar lead

4.2 Bipolar Lead (Fig. 1.2)

The bipolar lead method involves tracing the two active electrodes on the scalp connected to the G1 and G2 electrodes of the EEG machine without using extraneous electrodes. What is recorded by the bipolar lead method is the potential difference between the two active electrodes. When a monopolar lead shows an abnormal wave in a certain part, it can be confirmed on the bipolar lead; that is, phase reversed (or tit-for-tat) of the abnormal wave can be seen at the location where the abnormality occurs. The advantage of the bipolar lead method is that it is less susceptible to the influence of other bioelectricity, such as ECG, than monopolar lead, and it can eliminate artifacts caused by the activation of irrelevant electrodes. Bipolar leads must be combined and used with monopolar leads. Monopolar leads are the basis for analyzing EEG. Bipolar leads should be combined with the specific analysis of monopolar leads to draw correct conclusions. The bipolar lead design should be based on the principles of crisscross and triangulation [2, 3].

Fig. 1.2

Bipolar lead

4.3 The Parameters to Check

Electrode Impedance

After the electrode is installed, the impedance between the electrode and the scalp should be measured, and it generally should not exceed 5 kΩ. When possible artifacts caused by the electrode appear during recording, the electrode impedance should be checked again.

Calibration Voltage (Calibration)

Square wave calibration and biological calibration are required before recording. During square wave calibration, it is recommended to try to record and measure the calibration voltage under different filter settings. The calibration voltage should be adjusted to a sensitive level, and all the recording pen tips should be at the zero position and arranged in the same straight line. After biological calibration, the curves of each lead should be completely consistent in waveform, amplitude, and phase.

Sensitivity

In routine recording, the sensitivity is generally set at 7 μV/mm or 10 μV/mm (adults), 10 μV/mm, or 20 μV/mm (children). The sensitivity can be adjusted in time as appropriate.

Filtering

In a normal recording, the high-frequency filter should not be lower than 70 Hz and should be set to 70 Hz. The low-frequency filter should not be higher than 1 Hz and should be set to 0.3 Hz or 0.5 Hz (the corresponding time constant is 0.4 s or 0.3 s, respectively).

Paper Speed

The speed of normal paper feeding is set to 3 cm/s. For long-term tracing, 1.5 cm/s speed can be used.

Tracing Time

Routine EEG recording should last at least 20 min without interference in the awake state.

Induction Test

Eye opening and closing, flash stimulation, and hyperventilation should be used as routine induction tests, and sleep induction should be performed as much as possible. When conducting an induction test, it is necessary to increase the recording time accordingly.

• Eyes open and closed test: When the subject is awake and relaxed with eyes closed, he is asked to open his eyes for 3–5 s every 10s and to repeatedly open and close his eyes 2–3 times. The time of each eye opening and closing point is marked.

• Flash stimulation: The light stimulator is placed approximately 30 cm in front of the subject’s eyes, and his eyes are directed to look at the center of the stimulator with his eyes closed. The brightness of the stimulator is 100,000 candles (>100 Nit), the stimulation pulse width is 0.1–10 ms, and the stimulation frequency is adjustable between 1 and 60 Hz. The duration of each frequency stimulation is 10 s, and the interval is 10s. Then, another frequency stimulation is used for 10 s. Generally, the stimulation is gradually increased from low frequency to high frequency; for example, 1 Hz → 3 Hz → 6 Hz → 9 Hz → 12 Hz → 15 Hz → 18 Hz → 21 Hz → 24 Hz → 27 Hz → 30 Hz.

• Hyperventilation: Hyperventilation recording should last at least 3 min, and the deep breathing rate should be 20–25 breaths/min. Before and after hyperventilation, recording should last for at least 1 min without changing the lead combination. Hyperventilation should not be performed on patients with severe cardiopulmonary disease, cerebrovascular disease, high intracranial pressure (ICP), sickle cell anemia, or patients with poor general conditions.

• Sleep induction: The process of falling asleep and the light sleep period (nonrapid eye movement (NREM) I and II), deep sleep period and rapid eye movement (REM) sleep period should be recorded.

The lead combination methods commonly used in China are shown in Tables 1.1 and 1.2.

Table 1.1 Monopolar lead combination method

Table 1.2 Bipolar lead combination method

5 Interpretation of the EEG of Healthy Adults

Whether an EEG is normal is only a statistical concept. EEG analysis has a certain degree of subjectivity and relativity, and proper training and experience are very important. An EEG reflects the changes in the electrical potential of neurons. Therefore, any two diseases involving the same degree of neuron function will produce the same EEG abnormality; in contrast, an EEG abnormality can have multiple causes; therefore, the cause of a disease cannot be determined by EEG.

In the process of brain maturation, an EEG, which reflects the functional state of the brain, undergoes a developmental process from immaturity to maturity similar to that of the functional state of the brain. Therefore, individuals of different age groups have different EEG recordings; there can be multiple types of normal EEG between individuals of the same age. Moreover, the EEG described in the awake state is completely different from that described in the sleep state, and there are more inter-individual differences in awake–state EEG than in sleep-state EEG. At different stages from infancy to 19 years old, normal changes in EEG are very obvious. As age increases, EEG shifts to a more mature type. There is little change in normal EEG between 20 and 60 years old, the boundary between normal and abnormal is easy to define, and the types of abnormalities are more consistent with the basic types of encephalopathy. The EEGs of individuals older than 60 years old are similar to those of adults younger than 60 years old.

The EEG of a healthy awake adult is composed of α wave as the basic wave and a small amount of scattered fast and slow waves. Basic wave: mainly α waves or θ waves with normal distribution; symmetrical on both sides, the frequency difference of the α waves should not exceed 20% in the symmetrical part, and the amplitude difference should not exceed 50% in the occiput or 20% in the other part; the volatility should not be too high, and the average volatility of α waves is less than 100 μV. During eye opening and closing, mental activity, and stimulation sensation, α wave should have a normal response. Slow waves: These waves are scattered and of low amplitude, are mainly seen in the temporal area, and are mostly θ waves; there should be no continuous high-amplitude θ or δ waves in any part. Brain waves should be symmetrical during sleep. No abnormal electrical activity should be observed. Regardless of wakefulness or sleep, there should be no spike wave, spike-slow wave complex, etc.

The determination of an abnormal EEG is not based on a lack of normal EEG components or types but on whether the EEG contains abnormal activity or types. If an EEG contains abnormal electrical activity, no matter how many normal components it contains, it should be considered abnormal. In most abnormal EEGs, abnormal types do not completely replace normal electrical activity. They may appear intermittently or only in a certain area or areas or add to the normal background.

Abnormal EEGs are divided into four basic types: (a) epileptic activity, (b) slow wave, (c) abnormal amplitude, and (d) deviation from the normal type. Each type of abnormality may be caused by one or several types of brain diseases. Brain abnormalities are characterized by irritating or destructive lesions, which are located in the cortex, subcortex, or outside the cortex. On the other hand, many diseases cause more than one type of EEG abnormality, and for a neurological disease, not all cases have EEG abnormalities. If the brain lesions are small and long and located deep in the brain, the EEG may be normal. Although some people have abnormal EEG, they do not have any other manifestations of brain diseases. For the above reasons, EEG cannot be used alone for specific clinical diagnosis; it can only suggest a series of possible diagnoses. As with the results of other laboratory tests, EEG may be valuable in differential diagnosis and guide the choice of correct diagnosis. For example, in a patient with an unknown history of coma, the appearance of fast waves in an EEG may indicate barbiturate poisoning, the appearance of bilateral synchronous universal triphasic waves (TWs) is beneficial for the diagnosis of hepatic encephalopathy, and the appearance of focal slow waves or amplitude suppression may be beneficial for the diagnosis of subdural hematoma.

The following are the criteria for judging abnormalities in an adult EEG:

• The average amplitude of the basic rhythm is extremely high or flat and mixed in by slow waves of low amplitude.

• The basic rhythm lacks a response to various physiological stimuli on one or both sides.

• The basic rhythm wave amplitude is obviously asymmetric, >50%, or the frequency difference between the two sides of the wave is 20%.

• Slow wave activity exceeds the normal amount, especially when it is focal.

• There is a certain combination of spike waves, sharp waves, spike-slow complex waves, and sharp-slow complex waves in awake and sleep tracing.

• Slow waves and fast waves with high amplitudes erupt. More than two explosive activities occur during hyperventilation.

• The apical sharp wave, sleep spindle, and K composite wave that appear during sleep are obviously asymmetric.

6 Recognition of Common Artifacts in the EEGs of Patients in Severe Condition

An EEG is a weak bioelectric signal that needs to be magnified millions of times for the recording of this kind of electrical activity. In this process, non-EEG signals from various sources may also enter the amplifier and be mixed into EEG signals, which are called EEG artifacts. In the intensive care unit (ICU), a large number of artifacts are often produced by various instruments and equipment and the operations of staff. These artifacts are complex and diverse, and they often cause difficulties in the reading, analysis, and judgment of EEGs, thereby affecting the accuracy of EEG diagnostic results. Recognizing these artifacts is the key to improving the success of EEG diagnosis. EEG technicians often hope that the anti-interference ability of the instrument is as strong as possible, but the anti-interference ability of the instrument contradicts the fidelity of the EEG signal. Many interference signals are stronger than the EEG signal. The elimination of interfering artifacts may also cause some EEG signals to be lost, leading to the distortion of EEG waveforms. Therefore, some artifacts are inevitable, and the most important thing is to be able to identify them correctly and avoid making incorrect conclusions.

There are many artifacts in the ICU that need to be judged in real time based on the on-site observation of the patient’s condition. As time passes, it is difficult to determine the nature and source of the suspicious waveform. Therefore, in the identification of artifacts, the patient’s environment and ongoing activities need to be taken into consideration as much as possible, and it is also important to be familiar with the characteristics of the instrument. To improve the identification and elimination of artifacts, reduce the false-positive detection rate of EEG, and improve the accuracy of diagnosis via EEG, researchers need to understand the scenarios in EEG recordings and patient activities in real time or through video recording to help eliminate false positives.

Common artifacts in the ICU mainly include physiological artifacts, artifacts from instrument electrodes, and artifacts from environmental electromagnetic interference. Some common artifacts that occur in the ICU are summarized as follows.

6.1 Physiological Artifacts

Artifacts Related to Eye Movement

In patients with emotional stress or habitual blinking, blinking artifacts of 2–4 Hz continuous high amplitude consistent with the blinking frequency may appear in the frontal and temporal region, similar to the regular “θ wave,” especially in the prefrontal region (Fig. 1.3). In patients with eye movement, large slow waves can appear in the frontal and temporal regions. At this time, we can observe the patient’s eye movement or make his/her eyes open to record and compare them. If the artifact is due to blinking, gently pressing the eyelid with the thumb and index finger can eliminate the artifact. Soothing the patient by advising them not to be nervous and not to move the eyes can alleviate or eliminate eye movement artifacts.

Fig. 1.3

Eye movement artifacts

Electromyography (EMG) Artifacts

In patients with severe disease, abnormal mental behavior caused by disease factors or in the ICU, such as anxiety, tension, and dysphoria, can cause many EMG activity artifacts in EEG. The source can be the EMG potential, the friction potential of the clothes, the movement of the electrode wires, or the mutual contact of the wires. The graphics can be different due to the source of the artifacts, which are manifested by the baseline drift of the poor electrode contact or the single sharp pulse or continuous upper frequency sharp pulse. The waveform is indistinguishable when the upper frequency is released, and it is a thick line similar to a spicule (Figs. 1.4, 1.5 and 1.6). At this time, the operator should try not to use an upper frequency filter to eliminate the artifact but should ask the patient to relax, adjust their position, and distract their attention to reduce EMG interference.

Fig. 1.4

Artifacts caused by a patient repeatedly tapping the bed rail due to abnormal mental behavior

Fig. 1.5

The patient could not relax during EEG examination, and muscle tension is seen

Fig. 1.6

EEG showing EMG artifacts in a tetanus patient with whole-body muscle tension

Masticatory-Related Artifacts

When masticatory-related artifacts occur, there are 5–10 continuous clusters of high-amplitude irregular myoelectric bursts in the bilateral anterior and middle temporal leads (F7, F8, T3, T4). There is an obvious positive deflective slow wave at the end. The slow wave can sometimes spread to a larger range or even all leads, lasting 0.2–0.5 s, with an interval of 0.5–2 s, depending on the patient’s masticatory frequency while eating (Fig. 1.7). Masticatory interference is commonly observed in patients in ICU, such as those with abnormal mental behavior or in patients with autoimmune encephalitis (AE) with involuntary oropharyngeal movements. This pattern can easily be mistaken for spiny slow waves, and it is basically impossible to analyze during mastication. Therefore, patients who can cooperate should be advised not to eat during the examination, and patients who cannot cooperate should be monitored for as long as possible to extend the recording time.

Fig. 1.7

Artifacts caused by patient swallowing

ECG Artifacts

An artifact waveform that is consistent with the heart rate can appear in a single lead, a lead on one side and even all leads. This waveform can be a negative, positive, or two-way sharp ECG artifact, which is equivalent to the R wave of an ECG. There is no fixed relationship with background EEG activity, but the artifact is related to body position (Fig. 1.8). ECG artifacts are common in patients with short necks, obesity, and hypertension and when the ear electrodes are not properly positioned. In brain death monitoring, background brain electrical activity is almost equipotent. It is necessary to increase the sensitivity of the amplitude display (2 μv/mm), and it is difficult to eliminate the highly amplified tiny ECG artifacts. At this time, the patient’s body position can be changed, trying to keep it as straight as possible to extend the distance between the neck and the heart. ECG artifacts can also be eliminated by adding dual leads or average leads; adding another lead to record the ECG can quickly determine whether an artifact is an ECG artifact.

Fig. 1.8

In the judgment of brain death, artifacts consistent with heart rate appear on all electrically silent leads

Perspiration-Related Artifacts

Perspiration can cause changes in skin resistance, leading to a very slow baseline drift, but there is still brain electrical activity on an EEG (Fig. 1.9). If the baseline drifts up and down too much and exceeds the display range, the EEG cannot be analyzed. Perspiration artifacts often appear for a period of time after falling asleep. Although a patient may have no obvious perspiration, there are still changes in skin resistance and obvious baseline drift. To reduce this interference, the patient can be slightly awakened and then allowed to fall back asleep. Alternatively, a low-frequency filter can attenuate the slow baseline drift. Some patients with severe neurological illnesses perspire more, for example, in a patient with perspiration induced by stress. Resistance can be changed or electrodes can be loosened when hyperthermia causes profuse perspiration, resulting in very slow and unstable baseline movement with high amplitude, which can sometimes be masked by brain waves. At this time, the proper temperature can be maintained by physical cooling or adjusting the room temperature.

Fig. 1.9

Skin resistance changes during perspiration, causing baseline drift

6.2 Artifacts from Instruments and Electrodes

Artifacts Produced by a Respirator

Patients on a respirator move passively, and the respirator causes the electrode to shift and the electrode line to swing, which can cause a large slow wave such as a slow baseline drift that is consistent with the respiratory frequency. In this situation, attempts to balance the patient’s body and the removal and stabilization of the electrode wire attached to the shoulder and neck can be helpful to reduce artifacts (Figs. 1.10 and 1.11).

Fig. 1.10

Respiration artifacts consistent with respiratory frequency

Fig. 1.11

After adjusting the patient’s position, the respirator artifacts disappeared

Electrode Artifacts

Factors such as poor electrode fixation, long-term recording, dry conductive fluid, excessive scalp grease, and too much resistance between the electrode and the skin can cause electrode artifacts, which are intermittent or continuous irregular baseline drifts, accompanied by relatively many 50 Hz interferences, and occasionally a small number of interference waveforms, which are very similar to spike and sharp waves (Figs. 1.12 and 1.13). Therefore, if the electrode contact of a lead is not appropriate and a large number of disorderly interference waveforms often appear, extreme caution should be used when identifying abnormal EEG waveforms in this lead, so that the interfering wave is not misinterpreted as an abnormal EEG. Because procedures are frequently performed in the ICU, electrode loosening or falling off often occurs. Therefore, it is necessary to strengthen and fix the electrode and check whether the electrode is secure in an ICU.

Fig. 1.12

Artifacts caused by poor electrode fixation

Fig. 1.13

Artifacts caused by poor electrode contact

6.3 Artifacts from Environmental Electromagnetic Interference

AC Interference

The 50 Hz AC from various power sources and electrical appliances can easily interfere with EEG recording through electrostatic induction or electromagnetic induction. When various electrical equipment is started, the passage of electric current can generate a magnetic field around it. If the electrode or electrode wire of a patient’s head is within the range of the magnetic field line, electromagnetic induction can be generated, causing 50 Hz interference in the EEG recording, manifested as a 50 Hz wave appearing in one, several or all leads, superimposed on brain electrical activity. If the 50 Hz interference wave has a high amplitude, brain electrical activity will be completely covered (Figs. 1.14, 1.15, and 1.16). In the ICU, various medical equipment devices, such as monitors, respirators, sputum suction devices, and infusion pumps, are present around the patients. In this environment, EEG recording is very difficult. The 50 Hz interfering wave can be reduced by a 50 Hz notch wave during EEG recording or analysis, keeping the various instruments away from the EEG amplifier as much as possible. If necessary, some instruments can be turned off for a short time. If the patient’s body is connected to multiple electronic instruments, only one instrument earth wire is required.

Fig. 1.14

When using extracorporeal membrane oxygenation (ECMO), many AC artifacts will appear

Fig. 1.15

When using continuous renal replacement therapy (CRRT), many AC artifacts are seen

Fig. 1.16

After stopping ECMO and CRRT, the AC artifacts disappeared

Electrostatic Interference

Electrostatic interference causes a group of high-amplitude and irregular chaotic waveforms on an EEG, some of which are similar to spikes, sharp waves, or spike-slow complex waves, which should be distinguished (Fig. 1.17). To prevent electrostatic interference, patients should avoid wearing chemical-fiber clothing, and other people should avoid walking frequently beside the patient’s bed during the examination. Liquid dripping from an intravenous infusion set near the electrode line of the patient’s head in the ICU will also result in electrostatic interference, causing EEG artifacts (Fig. 1.18). At this time, the infusion time should be marked on the EEG. When the patient’s condition permits, the infusion set can be temporarily turned off, and this artifact can disappear.

Fig. 1.17

Electrostatic interference caused by frequent walking in the monitoring room

Fig. 1.18

Because the intravenous infusion set is located near the electrode line, electrostatic pulse interference is generated when the liquid drips

Case

In the patient Huang, with fulminant myocarditis, after cardiopulmonary resuscitation (CPR), the EEG showed full electrical inhibition. During the examination, the patient had sudden ventricular fibrillation. The EEG showed ventricular fibrillation patterns consistent with the cardiac rhythm, which showed a slow-wave-like pattern and rhythmic distribution (Figs. 1.19, 1.20, 1.21, and 1.22).

Fig. 1.19

The EEG background is in a state of full electrical suppression

Fig. 1.20

As the heart rate increases, the EEG gradually shows slow-wave-like artifacts consistent with the heart rate

Fig. 1.21

Slow-wave-like artifacts gradually become apparent on the EEG

Fig. 1.22

Slow-wave-like sample covering all leads