The EEG technician should have prior training on electrode placement and polygraphy of newborns to obtain a useful and technically satisfactory recording.[2] The technologist should be familiar with the artifacts in newborns and should have the knowledge to troubleshoot with minimal disruption of the running of a neonatal intensive care unit. At times, there may be difficulty in adequately accessing the newborn because of concurrent non-neurological monitoring, over-crowding related to equipment and hardware of ventilator, and ECMO procedures. The technician should work with the neonatal nurses, physician team, respiratory therapists, etc. to obtain a satisfactory recording without compromising the care of the infant.

During a neonatal recording, it is essential to pay attention to the following[3]

1. Minimize disturbance to the neonate.
2. Appropriate equipment preparation and positioning enables the technologist to observe and document gestational age (GA), conceptional age (CA), state of the infant, clinical data.
3. Minimize interference and artifacts. 
4. As the skin resistance in neonates tends to be high, skin preparation is of utmost importance. Adequate use of conducting paste helps in keeping the impedance below 5-10 kOhms.

Technical Considerations for EEG recordings in neonates[4]

1. A sampling frequency of 256 Hz is the recommendation. Screen display settings include a gain of 10µV/mm, a minimum time constant of 0.3 sec, filter settings (high pass 0.5Hz, low pass at 70 Hz).
2. Typically, the duration of the recording in neonates should include at least one complete sleep cycle (about 60-70 min ideally). However, under clinical situations, contingencies may force a shorter duration of the recording.
3. Paper speed may be set at the standard speed of 30mm/sec. We may sometimes use slower paper speeds because it compresses the record and facilitates visual recognition of delta activity, which is the dominant frequency seen in neonates.
4. A polygraphic recording should include at least eight scalp EEG electrodes, EKG, and respiration.
5. We use a modified 10-20 EEG electrode placement system permitting a combination of a longitudinal and transverse bipolar montage.

Long-term monitoring is avoided in newborns, especially in very premature infants due to:

1. Skin maceration

Skin breakdown and maceration can occur because of the electrode adhesives and traction on the electrodes. 

2. Chance of infections

Newborn infants are prone to infections due to their delicate scalp, and it is imperative to avoid contamination and placing leads in the proximity of any scalp lacerations and or sutures.

Some studies have shown to have better outcomes with dry electrodes.[5] 

Recognition of temporal maturational changes with age[6][7]

 I) Premature infant to the term infant:

General patterns of background activity change with gestational age and state. There are observable changes in amplitude, dominant frequency, periods of discontinuity, the synchronicity of bursts in extreme premature progressing to greater degrees of continuity, shortening of inter-burst intervals, fragmentation, and lability through 24-30 weeks. With the appearance of observable eye movements and changes in background activity, sleep state differentiation can begin at 25 weeks and completed by 30 weeks.

 II) Organization of behavioral states and developmental milestones in infants and children:

State differentiation is an important aspect of the neurophysiological assessment of neonates. The waking state includes two different states; one associated with agitation (active wakefulness), or without agitation (quiet wakefulness). We can also distinguish the sleep state into active sleep (associated with REMs) and quiet sleep (without REMs or non-REM/NREM). Beyond 35 weeks CA, active sleep can be seen before the onset of QS (AS1) and after QS (AS2). Between the two identifiable states above, there can be states with discordant features termed indeterminate or transitional sleep. The organization of activity changes described below is with respect to gestational age.

A. 24-25 weeks gestation to 28 weeks

EEG changes show an inconsistent correlation with changes in the state that usually alternate between periods of activity and rest. Background activity is markedly discontinuous; short duration runs of monophasic or diphasic delta activity (0.3-1 Hz) with superimposed theta rhythms. Delta activity of high amplitude (up to 300 µV) may be regional in expression occurring over temporal, occipital, and central regions can be bilateral or unilateral, while frontal delta activity is less frequent. Theta bursts may predominate in temporal areas, bilateral in expression, and become more abundant with progression to 28 weeks GA. 

B. From 28-29 weeks until 31 weeks

Behavioral states including active wakefulness (shows artifacts), quiet sleep (discontinuous), and active sleep with rapid eye movements (continuous or semi-continuous) become better defined progressively on the EEG tracing. Background activity shows periods of continuity up to 160 secs, with inter-burst intervals up to 30 sec. Delta activity shows a reduction in amplitude with increasing GA with superimposed theta or alpha rhythms. Delta waves are less diffuse, become more regionalized over occipital and central regions, while theta rhythms can occur in synchronized bursts or may undergo localization in the occipital and temporal areas. A few delta brushes (delta waves with superimposed alpha or beta rhythms) appear around this stage. By 31 weeks, we see delta activity 0.7-2 Hz with amplitudes of up to 200 µV, and the amplitude of theta activity also diminishes to about 20 µV. Delta brushes become prominent and diffuse. While synchronous delta activity is more common in AS, theta activity over temporal regions becomes more prominent in QS at this stage. We also see EEG reactivity to stimuli along with an attenuation of the amplitude of background rhythms.

C. 32 weeks to 34 weeks

At this stage, the sleep states become even more distinct; while the background activity is continuous in wakefulness, and AS, it is noticeably discontinuous in QS with an increase in burst duration between 32 to 34 weeks GA and a reduction in inter-burst intervals to < 15 sec at 32 weeks, and < 10 sec at 34 weeks. From 32 weeks, delta burst activity increases in frequency (1-2 Hz), but shows a reduction in amplitude, becoming exuberant or profuse by 34 weeks. Delta brush activity remains localized to occipital and 34 weeks GA. Theta activity seen in earlier GA disappears in AS at 32 weeks and from the QS by 33-34 weeks. Immature poorly defined frontal sharp transients may show up by 34 weeks GA.

D. 35 weeks to 36 weeks

By this stage, the waking and sleep stages show further differentiation and can be distinguished easily. In wakefulness, background activity is continuous with mixed polyfrequency (activité moyenne) activity, and active sleep (AS1 or first REM) stage precedes quiet sleep.

Here the activity is continuous and high amplitude slow activity, bursts of monomorphic activity (delta activity 1-3 Hz, 50-100 µV in frontal regions) become clear. We term this activity as anterior slow dysrhythmia. Delta brushes are more frequent in AS1. There is a transition from AS1 to QS marked by the appearance of discontinuous background activity, periods of relative attenuation lasting <10 sec. The second REM or AS2 phase is marked by the appearance of continuous but lower amplitude waveforms in the background and a greater amount of theta waves. Background activity shows bi-synchronous activity in active sleep and is asynchronous in quiet sleep.

E. 37 weeks and beyond Term

During this stage, we can make a clear differentiation between waking and sleep states based on electrographic features. The background activity in waking shows activité moyenne, AS1 showing mixed frequencies, and higher amplitude than in AS2, discontinuity, and trace alternant are features seen in QS. More specific features include localized expression of delta brushes in the occipital regions with progressive rarity beyond 40 weeks. Frontal sharp transients and anterior slow dysrhythmia in AS1 becomes more clear closer to term. Rolandic theta waves in AS1, and interhemispheric synchrony in QS are well marked. By 44 weeks, other features replace the characteristic EEG findings of the different waking states in a term infant.

Neonatal EEG recording can be challenging for inexperienced EEG technician and also the untrained electroencephalographer because of its unique features, limitations and technical data. 

The neonatal team including the nurses, respiratory technologists, infusion specialists, radiographers, and neonatologists should work together to achieve a near optimal recording. However, this communication is often not perfect. The information about the age and gestational age should be conveyed to the EEG reader, to decide on the maturity of EEG.

The technologist should be able to adapt to the restrictive environment in protecting the newborn from infections and injury. Nurses should communicate with the technologist about the contraindications and any specific infant related information to the technologist to help decide on the adoption of changes in protocol or lead placements.[8]