A Portable EEG Triage Model for Reducing pPNES Misdiagnosis
Let’s consider a hypothetical scenario: a patient has been experiencing seizure-like episodes for years, going to the ER every time one occurs and taking the medications prescribed by doctors. Different medications are tried, side effects pile up, yet nothing improves. Perhaps the diagnosis was wrong from the beginning.
In this paper I want to explain a problem faced by approximately %40 of patients that came into the ER with seizure-like episodes; and could not get rid of it for years,trying medication after medication, facing side effects. First, I will examine the main factors contributing to misdiagnosis and then present multiple potential solutions.
When a patient comes in with a seizure, having signs such as short-term confusion, jerking movements in the arms and legs, loss of consciousness, fear, anxiety, feeling of deja vu; doctors usually treat the patients with benzodiazepines (a treatment used for epilepsy by calming the electrical activity in the brain).These medications are highly effective for epilepsy because an epileptic seizure is caused by sudden bursts of abnormal electrical activity in the brain. It may start in one part of the brain spreading into various areas; affecting senses, emotions or movements; differentiating based on where the abnormal activity occurs. Epileptic seizures can also be detected through EEG monitoring, which records electrical activity in the brain during an episode; however the standard diagnosis relies on clinical observation alone, rather than EEG confirmation.
But among 980 patients aged 8 years or older diagnosed and treated for status epilepticus (in the sense of an epileptic seizure lasting 5 minutes or longer in this paper), 79 (%8.1) (Jungilligens et al., 2021) patients were discharged with a final diagnosis of pPNES. With that in mind there is also a range of research that points to bigger groups of people (up to %20-25) (Huff et al., 2024) getting misdiagnosed for epilepsy. I’m going to explore pPNES specially in this paper.
Prolonged psychogenic nonepileptic seizures (pPNES) are rooted in psychological or behavioral factors, not abnormal electrical activity. So treating with a medication that calms electrical activity has no positive effect on the patient. These episodes may last much longer than the characteristic short episodes, often leading into emergency or intensive care admissions.
Now there are a couple of reasons why this happens, the first being the similarity between the symptoms of the two conditions.
PPNES almost fully mimics epilepsy so while the patient is in a critical condition it is very high possibility that doctors will take the strongest precaution, benzodiazepines. But when that doesn’t work, if the patient is unfamiliar with their condition this occurrence may cause them additional stress and uncertainty. In such cases the standard way of diagnosis, clinical symptom observation, isn’t sufficient causing the doctors to use other diagnostic methods. The biggest difference between two cases lies in the EEG records; EEG can provide objective evidence of abnormal electrical activity associated with epileptic seizures, while PNES does not produce the characteristic epileptic electrical pattern during the event.
But the EEG machines are often not used, and this is not a “ medical mystery” the tool already exists, it is just not reaching the patients and my thesis is that pPNES misdiagnosis is, in part, a distribution problem disguised as a diagnostic problem.
Normally a full clinical video-EEG setup uses 20-32 electrodes, placed precisely across the whole scalp to address all parts of the brain correctly, high resolution amplifiers to clearly observe the areas and days of continuous recordings being read by a specialist. (Caprara et al., 2026) It is built to catch subtle localised abnormalities anywhere in the brain; and it is very useful for surgical planning and understanding subtypes of epilepsy. However, using this system requires: specialised equipment, trained neurophysiology technicians and multiday hospital stays. As a result these resources concantrate in major hospitals/urban centers, but if the patient falls out of that radius (lower income regions, developing healthcare systems etc.) they have a higher chance of getting misdiagnosed, structurally. Not necesarrily because their case is harder but because the appropriate tool isn’t accessible.
My solution is basically a lower cost EEG triage that is also more portable. This portable triage EEG device will include:
- Fewer electrodes (for example 6-8 instead of the standard 20-23) placed at the key spots, not the full scalp.
- Dry electrodes; a setup that does not need a scalp prep including conductive gel.
- An easier software that flags simpler binary outputs (Epileptic electrical activity: yes/no and the core details)
The disadvantages would be:
- Reduced spatial precision, meaning that the system may not be able to identify where in the brain the seizure originates.
- Reduced ability to catch rarer or subtler abnormalities the full setup would identify.
But the advantages would be:
- The device is cheap enough thanks to the lowered electrodes
- The setup and the preparation is simple enough to use in a general clinic or even a mobile unit thanks to dry electrodes and the eliminated need for wet preparation as well as the easier software.
- The device would be portable enough but also answer the one important question; Can we create a cheap first-line screening tool that identifies patients who require full neurological investigation?
These methods have already been tested in some studies individually. For example: a study in the emergency department compared a portable reduced-lead EEG against the full 23-electrode standard EEG on the same patients. The reduced version matched the full EEG’s findings in 100% of the 12 patients tested, correctly identifying nonconvulsive seizure activity in the one case where it was present. (Brenner et al., 2015)
Other studies pushed the electrode count lower down to 6-10 electrodes instead of the standard 21-23,that have been placed at key spots rather than across the whole scalp. A meta-analysis found that simplified 8-10 electrode setups reached around 75% sensitivity for seizure detection, which is a good number. (Caprara et al., 2026)
Researchers have built a low cost, dry contact EEG headset paired with a learning model, designed specifically to work off a reduced electrode montage and still the detections were accurate.(Wickramasinghe et al., 2025)
And the strongest device that matched the idea of a more portable device is the Ceribell. It is an FDA cleared device that uses 10 electrodes and 8 channel headband plus a recorder and cloud based review portal. It is not built for specialist neurophysiologists only; physicians, nurses, allied health workers can be trained to set it up themselves easily. One trial found the fast readout led to a 40% change in physician treatment decisions, which means it has prevented patients from misdiagnosis from the start. (Madakadze & McGill, 2023) Together, these findings suggest that reduced-montage EEG systems can be operated outside specialist settings while still providing clinically useful information for the operating healthcare workers.
But while the technological path is possible and clear, the main problem is accessibility. The epilepsy treatment gap sits around 10% in the US, but is 5-10 times higher in low and middle income countries. (Sen et al., 2025) In Latin America specifically, only 40-50% of people with epilepsy receive appropriate treatment and in rural areas that gap can reach 80-90%. (Leon-Rojas, 2026) Even in the US patients on Medicaid or Medicare or uninsured were significantly less likely to receive video EEG monitoring than higher income patients. (Miller et al., 2024) That means this isn’t just a global south problem, it’s in the wealthy healthcare systems too.
Lastly; if the hypothetical patient I started this essay with had the triage EEG monitor existing in the local clinic they were admitted to, they wouldn’t have suffered through various ineffective medications or unsolved diagnostic problems. Seizures of any kind ,eplieptic or non eplieptic, should be held as a serious matter, receiving the right tools and the methods to work through. So better accessibility and better conditions to set up for usage comes to a decision for us to deploy where it’s mostly needed.
Doa's Blog 
Yorum bırakın