Amplitude-integrated electroencephalography (aEEG) is a simplified form of brain wave monitoring used primarily in newborn intensive care units (NICUs). It takes the complex electrical signals of the brain and compresses them into a single, easy-to-read trend line that doctors and nurses can watch at the bedside for hours or even days at a time. While a standard EEG might produce pages of detailed squiggly lines that require a specialist to interpret, aEEG distills that information into a streamlined display that helps clinicians quickly spot seizures, track brain recovery, and make time-sensitive treatment decisions for critically ill newborns.
How aEEG Works
A standard EEG records raw electrical activity from the brain using many electrodes across the scalp. An aEEG starts with that same raw signal but then processes it through several steps to make it more compact and readable. The signal is filtered to focus on frequencies between 2 and 15 Hz, which captures the range most relevant to brain health in newborns. It is then amplified, smoothed, and compressed onto a display that uses a semi-logarithmic scale, meaning small voltages are stretched out and large voltages are compressed. This makes it easier to see meaningful changes at a glance.
The result is a continuous band on the monitor. The top edge of the band represents the highest brain wave amplitudes over a given period, and the bottom edge represents the lowest. Instead of scrolling through minutes of raw EEG tracings, a clinician can look at a single screen and see hours of brain activity trends condensed into one view. The signal is typically sampled 100 times per second, then averaged down into blocks, which further smooths the display and makes long-term patterns visible.
How the Electrodes Are Placed
One of the biggest practical advantages of aEEG is how few electrodes it requires. A single-channel setup uses just two scalp electrodes placed on opposite sides of the head at positions called P3 and P4 (roughly behind each ear, over the parietal regions), following the standard international system for electrode placement. A ground electrode is placed on the forehead to reduce electrical noise.
For a two-channel setup, two additional electrodes are placed at positions C3 and C4, closer to the top of the head. This second channel helps detect differences between the left and right sides of the brain, which is useful when doctors suspect a stroke or a seizure that starts in one specific area. Compared to a full EEG, which can use 20 or more electrodes and requires a trained technician to set up, aEEG is far simpler to apply, making it realistic for continuous monitoring in a busy NICU.
What aEEG Is Used For
The primary use of aEEG is monitoring critically ill newborns in the NICU. The most common reason for placing a baby on aEEG is hypoxic-ischemic encephalopathy (HIE), a type of brain injury caused by oxygen deprivation around the time of birth. HIE is also the most common cause of seizures in full-term newborns, and detecting those seizures quickly matters because they can worsen brain damage if left untreated.
Seizures in newborns are notoriously difficult to identify by observation alone. Many neonatal seizures produce no visible movements at all, while many normal newborn movements, especially in premature babies, can look like seizures but aren’t. This makes brain monitoring essential rather than optional. aEEG gives clinicians a continuous window into brain activity so they can distinguish true electrical seizures from normal movement and begin treatment promptly.
Beyond seizure detection, aEEG is used to monitor babies during and after therapeutic hypothermia (cooling treatment for HIE), to assess overall brain function in premature infants, and to watch for complications after brain infections or metabolic disorders.
Reading the aEEG Trace
Clinicians classify aEEG tracings into a few key background patterns based on the voltage of the upper and lower margins of the band:
- Continuous Normal Voltage: The upper margin stays above 10 microvolts and the lower margin stays above 5 microvolts. This is the healthy, expected pattern.
- Discontinuous Normal Voltage: The upper margin is still above 10 microvolts, but the lower margin drops below 5 microvolts. The tracing looks interrupted rather than smooth, and normal sleep-wake cycling is absent. This pattern suggests moderate abnormality.
- Burst Suppression: The trace shows stretches of very low activity punctuated by short bursts of higher voltage. This pattern indicates more significant brain dysfunction.
A healthy newborn’s aEEG also shows recognizable cycling between sleep states, with the band narrowing during active sleep and widening during quiet sleep. The presence or absence of this sleep-wake cycling is itself an important prognostic indicator. When normal cycling returns after a brain injury, it generally signals a more favorable outlook.
How Accurate Is aEEG?
Studies comparing aEEG to conventional EEG (the gold standard) show that aEEG detects seizures with a sensitivity of about 80%, meaning it catches roughly four out of five seizure events. Its specificity, or ability to correctly identify a non-seizure recording, sits around 50% in general use, though this improves to about 67% when the monitoring is specifically targeting babies with suspected seizures.
Where aEEG performs better is in assessing background brain activity rather than catching individual seizures. Detecting whether the overall brain pattern is continuous or disrupted reaches a sensitivity near 89% and a specificity around 55%. This is one reason aEEG is considered particularly valuable for prognosis: the background pattern over hours tells clinicians a great deal about how the brain is recovering.
Limitations Compared to Full EEG
The simplicity that makes aEEG practical also creates blind spots. Because it uses so few electrodes, focal seizures that occur in brain regions far from the electrode sites often go undetected. Short seizures can also be missed because the time-compressed display effectively averages them out. A seizure lasting only a few seconds may not leave a visible mark on a trace that compresses hours of data into centimeters of screen.
Conventional EEG, with its many electrodes and full spatial coverage, can track how a seizure moves across different brain regions. This spatial evolution is one of the key features neurologists use to characterize seizures, and it simply isn’t visible on aEEG. For these reasons, aEEG is best understood as a screening and trending tool rather than a replacement for full EEG. When aEEG shows something concerning, or when clinical suspicion remains high despite a normal-looking aEEG trace, a full conventional EEG is typically the next step.
Why aEEG Matters for Prognosis
One of the most valuable roles of aEEG goes beyond seizure detection: it helps predict long-term outcomes. The background pattern on aEEG in the first hours and days after a brain injury correlates with neurodevelopmental outcomes months and years later. A baby whose aEEG trace returns to a continuous normal voltage pattern relatively quickly after birth tends to have a better neurological prognosis than one whose trace remains severely suppressed or shows persistent burst suppression.
This prognostic information directly shapes treatment decisions. For babies with HIE undergoing cooling therapy, the aEEG trace helps the medical team gauge whether the brain is responding and provides families with early, though not absolute, information about their child’s trajectory. It is one piece of a larger clinical picture, but it is often one of the earliest objective indicators available.