“Your EEG was negative.” Don’t bet on it.

A couple of weeks ago I mentioned at the end of my post that most physicians are limited in what they can offer a person experiencing visually induced seizures. If a doctor investigates possible seizures, an EEG (electroencephalogram) is usually the first diagnostic procedure. EEGs can (sometimes) detect problems with the electrical function in the brain.

If you’re told that an EEG “didn’t find anything,” don’t assume there was nothing to be found. Many issues can and do routinely interfere with the accuracy of EEG results, an open secret that clinicians rarely share with patients. In today’s post, I’ll present some of the shortcomings of EEG when used for seizure detection, as described by experts.

Let’s start with this:

“The main limitation with EEG is its poor sensitivity for epilepsy. The generally accepted numbers are that the yield of a single routine EEG in epilepsy is ~ 50% and increases with repeated EEG recordings to reach about 80% by the third recording.”

Source:  Selim R. Benbadis, “Introduction to Sleep Electroencephalography” in Sleep:  A Comprehensive Handbook, edited by T. Lee-Chiong, John Wiley & Sons, Inc., 2006

Translation:  EEG is more useful in assessing conditions such as coma, stroke, tumors, and brain death than it is in accurately detecting seizure activity. Half of routine EEGs don’t pick up seizure activity during the first recording. In cases proven to involve seizure activity, if the EEG is repeated a couple of times, 80 percent of them will capture evidence of seizures.


A negative result means…just that nothing was found. Whether or not an abnormality exists that might have been detected:

“While…often adequate for the appropriate diagnosis of a seizure disorder, patients with epilepsy can have persistently normal EEGs.”

Source:  Gregory A. Worrell, Terrence D. Lagerlund, and Jeffrey R. Buchhalter,“Role and Limitations of Routine and Ambulatory Scalp Electroencephalography in Diagnosing and Managing Seizures.” Mayo Clinic Proceedings September 2002**

Translation: A seizure disorder should be diagnosed on the basis of the clinician’s judgment, using information from the patient’s history, the clinician’s own observations/examination, and diagnostic tests. Doctors should not allow EEG results alone to drive the diagnosis, and patients should question a diagnosis that is based primarily or exclusively on the EEG.


 A bit of historical context:

“The human electroencephalogram (EEG) was discovered by the German psychiatrist, Hans Berger, in 1929…EEG continues to play a central role in diagnosis and management of patients with seizure disorders—in conjunction with the now remarkable variety of other diagnostic techniques developed over the last 30 or so years—because it is a convenient and relatively inexpensive way to demonstrate the physiological manifestations of abnormal cortical excitability that underlie epilepsy.”

Source: Shelagh J.M. Smith, “EEG in the diagnosis, classification, and management of patients with epilepsy.” Journal of Neurology, Neurosurgery, and Psychiatry June 2005 

Translation:  Folks, EEG  is pretty old technology. It was updated to incorporate digital technology, so that instead of recording the data on paper (think of seismographs), results are now viewed and stored on computers. However, the capture of electrical signals from inside the brain using electrodes on the scalp is much the same as it was in 1929. More specifically,

  1. There are other diagnostic procedures that have been developed more recently. (You may find it difficult to convince a neurologist to order one of these others, though, without having some evidence of seizures showing on the EEG. A real Catch-22.) 
  2. EEG is economical for providers and risk-free/non-invasive for patients.

Ready to get just a little technical? Remember, a translation follows:

“Despite the importance and widespread use of scalp EEG epileptiform discharges, the cortical EEG substrates underlying these spikes and seizure discharges are mostly speculative…the cortical area of epileptiform discharges required for the scalp recording is considerably larger than commonly thought. A cortical area of 10 to 20 cm is often required to generate a scalp recognizable interictal spike or ictal rhythm. Sufficient cortical source area and synchrony are mandatory factors for the corresponding scalp EEG epileptiform recording.”

Source:  J.X. Tao, M. Baldwin, S. Hawes-Ebersole, J.S. Ebersole, “Cortical substrates of scalp EEG epileptiform discharges.Journal of Clinical Neurophysiology April 2007

Translation: Data retrieved via EEG requires a lot of guesswork to figure out what is really going on inside the brain, and where.

  1. EEG electrodes glued onto the scalp do a so-so job of picking up the brainwaves that signal a seizure. For one thing, the skull gets in the way…for another, brain tissue.
  2. Scalp electrodes aren’t great at correctly identifying the location of abnormal discharges, either.
  3. Seizures involving only a small portion of the brain may not generate a strong enough electrical charge to pass through the skull for detection on the scalp.
  4. The spike and wave patterns on an EEG that characterize seizure discharges must arrive at the scalp undistorted by the intervening structures (many do not reach the scalp without distortion), or the waves will not clearly show seizure activity occurring.  

Are you developing some healthy skepticism?  Good!  Next up:  Using EEG to test for sensitivity to flash/flicker carries a whole set of additional challenges that further limit EEG’s usefulness.

**full text available here

Update, 7/26/2014:  Please read here about a new study that found only 6.2% of patients with a history of visually induced seizures showed a positive result on EEG during photic stimulation.


9 Comments on ““Your EEG was negative.” Don’t bet on it.”

  1. Marja Erwin says:

    While the eeg may be missing some people’s photosensitive epilepsy, there are other photosensitivities, too.

    I have a severe low-frequency photosensitivity, so flashing lights such as turn signals, hazard lights, crosswalk lights, police-car strobe-weapons, ambulance strobe-weapons, school bus strobe-weapons, etc. blind and disorient me, and if sustained, can trigger migraines, with vomiting. I had an eeg because when flashing lights come from multiple directions they sometimes cause me to stumble into the street.

  2. Hilary says:

    Thank you for the above. My daughter has just been told she has Non Epileptic Seizures as the EEG did not pick up any of the seizures she had whilst on the Ambulatory EEG.

  3. Michelle Dixon says:

    What your article fails to mention is that all of these issues that can adversely affect an EEG are all factored in. For every test that doesn’t pick up an actual seizure or epileptiform discharge, there is one that does. And while you may have done reading on the subject, until you actually know exactly how to read an EEG, factor in all the variants that affect it, and have the practical ability to hook up an EEG, your scope is very limited. Much in the way that a student who has read the books knows some things, but nothing compared to actually working in the field. And i would go even further in debating the diagnostic abilities of EEG, most especially in children. When working in a children’s hospital, i saw more abnormal EEG than normal.

    • jsolodar says:

      EEG can confirm the presence of epilepsy, but due to known limitations of the technology, this test should not be used to rule out seizures. That’s what the research says. Nonetheless, in practice, frequently it is used that way.

  4. Kari Strickland says:

    I believe that another problem with EEG is how the test is administered. I have had 3 EEG’s in the last 10 years and none have lasted more than 30 minutes. I also have very thick hair and each time they had trouble even getting the sensors to attach to my scalp and several sensors fell off during the tests. I don’t trust the results

    • jsolodar says:

      I totally agree. It makes no sense that the results from these very flawed procedures are routinely used as the deciding factor in clinical treatment.

      It is possible to get an outpatient, ambulatory EEG where the electrodes are hooked up for 24-48 hours. Not only do they get a longer sampling, but they may do a more careful job of gluing down the contacts. They bandage up your head more to better prevent them from loosening.

      Even better: dense array EEG, which offers up to 256 electrodes (instead of the typical 20 or so) over a much larger surface area of the head and neck. I only recently learned of it and don’t know how available it is outside of research settings.

  5. I admire the valuable data you provide about Electroencephalogram (EEG)
    in your blog.

  6. AEEG says:

    The biggest concern when using EEG for diagnosis of epilepsy or seizures is doing a long term EEG (Video or ambulatory). EEG is not so much inaccurate, unreliable or outdabted which i think this article seems to insinuate, it’s just that with seizures, your brain actually has to be having some sort of abnormal activity during the test, unlike a stroke where the damage is constantly there. If a patient is hooked up for a long term EEG and actually has 1 or a few of their “seizures” or events during the EEG, then there is little chance of it being incorrectly labeled as normal, or non epileptic in nature.

    • jsolodar says:

      Thanks for your comment. Because my daughter has experienced many seizures that were not picked up on EEG, I’ve done a lot of reading about this issue. In addition to what I’ve cited above, there are a number of other studies pointing out the limitations of scalp EEG. The limitations apply in particular to partial seizures. Here are a few more excerpts:

      1) “…despite an identical mechanism of seizure generation, epileptic seizures may often be EEG-negative (but seen only with MEG) because of their small magnitude.”

      Kakisaka et al., “Magnetoencephalography’s higher sensitivity to epileptic spikes may elucidate the profile of electroencephalographically negative epileptic seizures.” Epilepsy & Behavior Volume 23 Number 2 pp 171-173 (February 2012)

      2) “…mesial temporal seizure activity is commonly not recorded by scalp tracings.”

      Blumer, “Dysphoric disorders and paroxysmal affects: recognition and treatment of epilepsy-related psychiatric disorders.” Harvard Review of Psychiatry Volume 8 Number 1 pp. 8-17 (May-June 2000)

      3) “…Few, if any, epileptiform discharges confined to this structure [hippocampus] had a scalp EEG correlate. This includes hippocampal preictal spiking, the typical focal hippocampal seizure onset, and many “subclinical” hippocampal seizures. Likely reasons for this lack of identifiable scalp potentials are a relatively small activated tissue volume, curved geometry (which encourages external field cancellation), plus the usual shielding effect of the skull and scalp. Only when adjacent temporal cortex became involved in the seizure did scalp rhythms appear.”

      Pacia and Ebersole, “Intracranial EEG substrates of scalp ictal patterns in temporal lobe foci.” Epilepsia Volume 38 Number 6 pp. 642-654 (June 1997)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s