EEG in epilepsy

William O Tatum IV DO FACNS (

Dr. Tatum of Mayo Clinic received consulting fees from Bioserenity..

Nimit N Desai MBBS (

Dr. Desai of Mayo Clinic has no relevant financial relationships to disclose.

Jason Siegel MD (Dr. Siegel of Mayo Clinic has no relevant financial relationships to disclose.)
John M Stern MD, editor. (

Dr. Stern, Director of the Epilepsy Clinical Program at the University of California in Los Angeles, received honorariums from Greenwish, Sunovion, and UCB as an advisor and from Greenwich, Eisai, LivaNova, and UCB as a lecturer.

Originally released November 10, 2008; last updated February 23, 2020; expires February 23, 2023


Electroencephalogram is the most useful test when evaluating people with possible epilepsy. It may provide specific neurophysiological information to support the clinical diagnosis. Furthermore, based on the type of interictal epileptiform discharges, it may help classify the seizure type or epilepsy syndrome. EEG findings can guide management, from directing antiseizure drug management to localizing an epileptogenic zone for neurosurgical treatment. Furthermore, it is an important adjunct to the clinical examination in the critically ill patients for diagnosing and treating unrecognized seizures and nonconvulsive status epilepticus. The utility of EEG has extended from a widely available, versatile, portable electrophysiological study to a sophisticated computer-based clinical and research metric that is elemental in exploring fundamental brain function.

Key points


• EEG is the diagnostic test of choice when evaluating a patient with seizures and epilepsy.


• Epilepsy is a clinical diagnosis that is supported by interictal epileptiform discharges on the EEG and can be confirmed when EEG records an ictal rhythm during a seizure.


• EEG may help classify seizures and epilepsy syndromes by defining the distribution of epileptiform abnormalities and quantifying the frequency of seizure occurrence.


• Misdiagnosis of nonepileptic events as seizures is probably not rare due to EEG misinterpretation.


• The surgical treatment of epilepsy relies on ictal EEG to characterize the electroclinical localization of the epileptogenic zone.


• Special waveforms, recording conditions, and new techniques are expanding EEG usefulness beyond the application in epilepsy to other areas involved in brain disease.

Historical note and terminology

Historical note. In his seminal work “Uber das Elektroenkephalogram des Menschen” (“On the EEG of Man”), Hans Berger pioneered the discovery of human EEG, first recorded in 1929 (Haas 2003). The practical usefulness of EEG became apparent in the 1930s after interictal discharges were demonstrated first by Fisher and Lowenback and later by Gibbs, David, and Lennox in the United States.

In 1936, W Gray Walter demonstrated that this technology could aid in the diagnosis of tumors, stroke, and other focal brain disorders. For 40 years, EEG was the cornerstone to the diagnosis and treatment of seizures and epilepsy. Until the advent of CT and MRI, it was the first-line neurodiagnostic test for diagnosing tumors, stroke, and other focal brain disorders.

EEG data were analyzed by visual inspection until the 1960s. With the introduction of digital equipment in the 1960s and 1970s, the application of Fourier analysis to computer-based EEG algorithms, spectral analysis and quantitative EEG became a reality. Despite the innovation of neuroimaging, EEG continues to have a vital role in the evaluation of neurologic disease, expanding from use in seizures and epilepsy to also include encephalopathy, traumatic brain injury, sleep disorders, coma, and brain death.

Terminology. In 2017, after 35 years, the International League Against Epilepsy (ILAE) released a new classification of seizure types, largely based on the existing classification originally formulated in 1981 (Fisher et al 2017) and further standardized terminology for seizure classification. In broad terms, epilepsy syndromes are classified as generalized, focal, or unknown with an etiology of genetic, structural-metabolic, or unknown (Fisher et al 2017). Seizures are classified according to where they start in the brain. Generalized seizures involve both sides of the brain or large hemispheric networks of neurons at onset. Focal seizures (ie, partial seizures) refer to those seizures that start in one side of the brain and involve regional networks. Focal seizures may start on the surface of the brain or in deeper structures and remain restricted to an area or spread to involve larger network. Unknown seizures reflect those without a clear onset at the beginning or when atypical and mixed forms of focal and generalized seizures co-exist. The new terminology correlates to older language (Table 1), which is still often used clinically.

Table 1. Comparison of Old and New Terminology Set Forth by the ILAE in 2017


New terminology

Old terminology

Type of seizure and origin of onset


Originating within, and rapidly engaging, bilaterally distributed networks



Originating within networks limited to one hemisphere



Descriptors of focal seizures


Wording changes


Focal aware*

With observable motor or autonomic components

Simple partial


Involving subjective sensory or psych phenomena only


Focal impaired awareness


Complex partial dyscognitive

Focal to bilateral tonic-clonic seizure**


Secondarily generalized seizure

Seizure etiology


Epilepsy as a direct result of a known or presumed genetic defect in which seizures are the core symptom of the disorder



Distinct structural or metabolic condition or disease that substantially increases epilepsy risk



Nature of underlying cause is unknown

Unknown onset, seizure type


The application of the standard scalp EEG in epilepsy has relied on the electrocerebral activity largely between the 1 to 30 Hz bandwidth (ie, Berger's bandwidth). Filter settings are increasingly being “opened” during video epilepsy monitoring to obtain full-band EEG, providing a more comprehensive approach (Tatum 2014a). Newer applications of frequencies (ie, gamma and high-frequency oscillations) are now being explored to enhance disclosure of brain regions and the networks that are involved in seizure genesis (Table 2).

*Focal seizures are classified as motor and nonmotor. Adding descriptions of other signs and symptoms is suggested (ie, focal aware motor seizure).

**The term “bilateral” is used for tonic-clonic seizures that propagate to both hemispheres, and “generalized” is a term used for seizures that appear to begin simultaneously in both hemispheres.

Table 2. Bandwidth and Interpretation of EEG Waveform Frequencies

Frequency (Hz)




0.0 - 0.5

Infraslow activity*


Onset of focal seizures

0.5 - 3.5


Sleep, HV, PSWY, elderly

Encephalopathy, white matter lesion

>3.5 - <8.0


Drowsiness, children, elderly

Encephalopathy, white matter lesion

8 - 13


PDR, mu rhythm, “third” rhythm

Ictal rhythm in seizure, alpha coma

13 - 30


Medication, drowsiness

Breach rhythm, drug overdose, ictal rhythm

30 - 80


Voluntary motor movement, learning/memory


80 - 250


Cognitive processing/memory

Interictal and ictal seizure frequency, possible epileptogenesis

250 - 500

Fast ripples*


Focal seizures

500 - 1000

Very fast ripples*

Acquisition of sensory information


* = Expanded frequencies currently under investigation

The characteristic interictal EEG features of epilepsy are spikes (20 to 70 msec) and sharp waves (70 to 200 msec) when displayed on a review monitor at a display speed of 30 mm/second.

Epilepsy’s characteristic EEG features Image: Epilepsy’s characteristic EEG features
Pathological discharges are distinguished from the background and contain a sharp contour, physiological field, rapid rise, and brief duration, occasionally with an after-going slow wave. Spikes and sharp waves can present in isolation or as polyspikes or polysharp waves.
Juvenile absence epilepsy EEG features Image: Juvenile absence epilepsy EEG features
Spikes and sharp waves possess the same potential for seizure genesis independent of morphology.

Each type of EEG recording in epilepsy has its own advantages and disadvantages. Standard scalp EEG recordings, short-term EEG, computer-assisted ambulatory EEG (CAA-EEG), and in-patient continuous video EEG monitoring (VEM) are different methods of EEG recording (Leach et al 2006), each with different clinical benefits and limitations (Table 3).

Table 3. Comparison of EEG Methods Used in the Evaluation of Patients with Paroxysmal Neurologic Events





Epilepsy Monitoring

Time to complete
Yield of recording
Natural conditions
Cost of study

30 - 60 minutes

4 - 24 hours

24 - 48 hours
+ / ++
++ / +++
$$ / $$$
$$ / $$$

1 - 7 days

Notes: The “plus rating” and “dollar sign” are used to identified relative relationships. A 1-plus (“+”) rating has the lowest and 3-plus (“+++”)the highest association with the feature. * Short-term EEG is based on a mean of 4 hours of recording. CAA-EEG = computer-assisted ambulatory EEG.

Scalp EEG is the simplest, least expensive, most practical, and, therefore, most common method to acquire EEG during routine clinical use. A standard scalp EEG is usually a brief 20- to 30-minute recording (but may extend up to 60 minutes duration). It is, therefore, typically inadequate for capturing infrequent paroxysmal neurologic events and abnormalities (Benbadis et al 2004). A prospective study of EEGs from 1803 patients revealed that only 19% of the patients had interictal epileptiform discharges in the first 30 minutes of recording, but the event capture rate increased by 30% during longer recordings (Burkholder et al 2016). However, increase in diagnostic yield with longer duration is not supported by other studies (Mahuwala et al 2019). Prolonged EEG is better able to capture seizures and neurologic events. One 5-year study of 175 outpatient short-term EEGs (shorter than 24 hours) found that 7% yielded seizures during the recording (Seneviratne et al 2012b).

A higher diagnostic yield (40%) was found in children when EEG monitoring was longer than 6 hours (Srikumar et al 2000). Prolonged the EEG recording either using CAA-EEG or inpatient video EEG monitoring offers distinct advantages. The yield of identifying epileptiform discharges to support a clinical diagnosis of epilepsy has been approximately 2.0 to 2.5 times that of a standard EEG and was cost-effective when compared to the gold standard of video EEG monitoring with a yield of ambulatory-EEG greater than 70% (Dash et al 2012).

The content you are trying to view is available only to logged in, current MedLink Neurology subscribers.

If you are a subscriber, please log in.

If you are a former subscriber or have registered before, please log in first and then click select a Service Plan or contact Subscriber Services. Site license users, click the Site License Acces link on the Homepage at an authorized computer.

If you have never registered before, click Learn More about MedLink Neurology  or view available Service Plans.

Find out how you can join MedLink Neurology