Epilepsy and EEG seizure-detection

Epilepsy, a chronic neurological disorder affecting approximately 50 million individuals worldwide, primarily manifests through seizures—abnormal electrical signals within the brain. The electroencephalogram (EEG), a non-invasive brain recording technique, serves as a precise tool in detecting these seizures. In this post, we'll explore the distinct characteristics of seizure-associated signals and the essential diagnostic procedures for understanding and managing epilepsy.

​What is an EEG and what does it measure?

The Electroencephalogram (EEG) is a neuroimage technique that uses an EEG machine to record the electrical activity generated when groups of neurons communicate. This process involves placing multiple EEG electrodes  attached the scalp surface. The resulting electrical activity is then amplified and displayed as wave signatures on a computer screen.

One of the primary advantages of EEG technologies is their exceptional temporal resolution, enabling the capture of neurocognitive processes occurring within tens to hundreds of milliseconds.

Over time, neuroscience research has consistently showcased the correlation between EEG signals and cognition. This has led to the establishment of well-accepted theories and empirical data supporting the relationship between brain function and neural signals (Berger, 1929; Gevins and Schaffer, 1980).

EEG unravels fascinating insights, such as pinpointing the specific brain areas activated during mental activities and detecting various neurological disorders like epilepsy syndrome or seizure disorder (e.g. Adebimpe et al., 2016). Indeed, one of EEG's most valuable capabilities is its precision in identifying abnormalities in brain waves, offering unique insights into a patient's level of impairment.

There are many different types of EEG systems and techniques. In a conventional EEG, an electrode montage is attached to an elastic cap according to the 10/20 system landmarks (Jasper, 1958). Today, we can find EEG systems that also provide a fixed electrode layout optimized for specific applications. 

In terms of sensor types,there are gel, water-based, metal-based dry EEG electrodes and even textile EEG electrodes. The choice between them hinges on the conductive substance required to reduce skin impedance and enhance signal quality. While gel-based EEG electrodes dominated research and medical settings in the past century, today's market offers more advanced systems utilizing water-based and dry EEG sensors.

These innovative EEG systems have emerged as alternatives to gel, offering enhanced user comfort and simplified setups for researchers, all while maintaining signal quality (Kam et al., 2019). For a performance comparison of gel, water (semi-dry), and dry EEG systems within the realm of brain-computer interfaces, refer to the study (Schwartz et al., 2020). 

How can EEG aid in Epilepsy?

A primary function of EEG is to aid clinicians and neurologists in accurately diagnosing epilepsy (Alebesque et al., 2017; NICE, 2012).  

As epileptic seizures involve atypical patterns of electrical discharges in the brain, known as epileptiform signals, EEG tests play a crucial role in seizure detection and classification.  

Furthermore, EEG's high specificity (i.e., false-positive detection occurs in only 0.5-3.5% of cases, WHO, 2012) is clinically significant, especially in cases where seizure manifestations are unclear (e.g., non-convulsive epileptic states) or when a differential diagnosis is necessary.

EEG plays a vital role in providing medical guidance and managing epilepsy. The unpredictable frequency and intervals between seizures (interictal periods) often pose challenges in controlling a patient's epileptic status. In this context, EEG serves to predict or assess the risk of seizure relapses and recurrence. This predictive capability is particularly valuable in guiding medical treatment decisions and evaluating their effectiveness.

Seizures arise from abnormalities in brain waves, and EEG serves as a powerful tool capable of offering unique insights into the landscape of a patient's epilepsy. Moreover, EEG is instrumental in predicting and monitoring ongoing seizure activity, providing invaluable information for effective management and treatment.

Furthermore, EEG monitoring is crucial for controlling ongoing seizure activity, which may be subtle or even absent, particularly during medication intake or after withdrawal (Berg and Shinnar, 1994; Smith, 2005).

One important consideration in EEG usage is the necessity to mitigate sources of errors that could impact the accurate management of patients with seizures. Potential error sources include subpar technical equipment quality and unskilled technicians. It is crucial to ensure a minimal technical qualification that meets the standards of ABRET-certified laboratories (e.g. American Clinical Neurophysiology Society Guidelines). 

For instance, a reliable EEG device must possess a high signal-to-noise ratio to differentiate epileptic spikes from background noise (Iwasaki et al.,2005). Following confirmation of EEG quality, test results should be interpreted by a specialized neurologist with in-depth knowledge of seizures and epileptic syndromes. Limited experience or lack of training in EEG interpretation is one of the most common reasons for incorrect diagnoses (Bendabis, 2010).

There are three primary applications where EEG proves highly valuable for epilepsy:

1. Medical diagnosis and management: EEG plays a pivotal role in aiding clinicians and neurologists to establish accurate diagnoses (Alebesque et al., 2017; NICE, 2012).     
    
2. Research: As a research tool, EEG has value for studying the onset, spread, and cessation of seizures. Questions concerning why certain seizure types emerge at specific ages or in response to particular events, the correlation with cognitive decline, and the effects of anti-seizure medications on brain cells are prominent areas of scientific investigation (Helmstaedter and Witt, 2017; Holmes, 2013; Rogawski, Löscher, and Rho, 2016)
    
3. Early detection: Certain EEG devices have the capability to alert patients or caregivers of impending seizures several minutes in advance.

What is detected in an EEG test?

In patients suspected of suffering from epilepsy, it is advised to perform an EEG test after the first seizure that is not known to be caused by alcohol/drug withdrawal or other medical conditions (e.g., encephalopathy; Pohlmann-Eden et al., 2006). Once the diagnosis is established, an EEG test will provide accurate information about ongoing brain electrical activity during the test. It will inform the doctor about when a seizure begins, how and when the seizure spreads, and when it stops.

Seizure EEG can be characterized according to delimited periods of time. For instance, the term ictal refers to the EEG activity recorded during a seizure event whereas interictal and postictal indicates the EEG activity captured between and after seizures respectively (Britton, Frey and Hopp, 2016). In the diagnosis of epilepsy and seizure onset localization, these recordings entail different wave morphologies resulting highly informative.

1. Ictal EEG patterns

For clinical purposes, seizures are classified into two main categories: focal-partial seizures and generalized seizures (Blume, 2010; Sharbrough, 1993).The key distinction between these types of seizures lies in their origin.

1. Focal Seizures: These seizures originate from a specific area of the brain. Typically, focal seizures begin in subcortical structures and are confined to one hemisphere. They usually last about one or two minutes and are commonly associated with Temporal Lobe Epilepsy. 
2. General Seizures: In contrast, generalized seizures involve a widespread electrical discharge affecting both brain hemispheres simultaneously.

During epileptic seizures, abnormal EEG signatures often manifest as rapid spiking waves (30-80ms) or sharp waves (70-200ms), which may or may not be followed by slow waves. (Fischer, 2014) outlines at least five distinct patterns of epileptic waveforms observed during an ictal EEG:

1. Rhythmical frequencies: These evolve in the delta (0–3Hz/s), theta (4–7Hz/s), or alpha (8–12Hz/s) bands with varying degrees of sharpness.
2. Rhythmic fast-spiking Seizures: Common in seizures with focal onset, such as those originating from the hippocampus. These seizures exhibit frequencies of 40-50Hz/s.
3. Spike-Wave Complexes: These are generalized discharges with frequencies ranging from 1.5 to 2.5Hz, featuring sharp peak durations of 20 to 70 milliseconds. Such epileptiform activity can occur during focal or generalized tonic-clonic seizures.
4. Electrodecremental Patterns: Reflecting seizures characteristic of temporal lobe epilepsy, these patterns are marked by a general flattening of brain rhythms at the onset of a seizure.
5. Clinical Seizure without EEG Changes: This occurs when there are no observable changes in EEG activity during a clinical seizure, suggesting a seizure origin distant from scalp electrodes.

2. Interictal and postictal EEG patterns

Interictal EEG provides insights into abnormal brain activity between seizure intervals. Patients with a history of epilepsy often exhibit a specific pattern of pathological activity known as interictal epileptiform discharges (IEDs) during this type of EEG test, distinct from the activity observed during seizures.

However, research indicates that not all epileptic patients display such obvious abnormal brain activity (Pillai and Sperling, 2006). Consequently, aside from seizure intervals, EEG activity may appear entirely normal during other times.

Following a seizure event, a recovery period called the postictal interval ensues (Abood and Bandyopadhyay, 2019). During this phase, the brain recuperates from the seizure's effects. Postictal EEG recordings typically reveal attenuation or slowing in slow-wave activity (delta-theta waves). The average time for EEG to return to baseline is approximately 2 hours, with a maximum of 8 hours in adults (Arkilo et al., 2013), depending partly on the seizure's severity, location in the brain, and type. Common symptoms during this period include migraine, drowsiness, confusion, and altered states of consciousness like memory loss.

What is the EEG procedure for epilepsy diagnosis?

Diagnosing epilepsy using EEG procedures helps provide a comprehensive understanding of a patient's seizures. Typically, the test is conducted by an EEG technician who records and monitors brain activity while the patient is awake or asleep.

Before the EEG test, the patient is prepared for recording, a process that typically lasts between 20 and 25 minutes. The EEG room should be quiet, well-conditioned, and often dimly lit.

During the EEG test, the technician records brain activity while asking the patient to report any seizure symptoms they may experience. Importantly, the EEG procedure is painless, comfortable, and generally carries no risk.

Different types of EEG tests may be employed, each with its own procedures:

• Routine EEG: Conducted in a specialized clinic or hospital, lasting less than 1.5 hours.
• Sleep EEG: Used when a routine EEG does not yield sufficient information, with procedures similar to awake EEG but conducted while the patient is sleeping.
• Prolonged EEG: Involves continuous monitoring in an epilepsy monitoring unit for 1 or 2 hours, or over several days as an inpatient. Video recording may be used concurrently to capture seizure onset and characteristics.
• Ambulatory EEG: Conducted in outpatient settings or the patient's home over several hours or days.

In situations where seizures may not occur during recording, epileptiform activity can be induced using standard seizure activation procedures. These include:

1. Rapid eye opening and closing.
2. Photic stimulation (e.g., exposure to flashing lights).
3. Rapid or deep breathing (hyperventilation).
4. Sleep deprivation (staying awake the night before the EEG test).

Wearable devices for epilepsy seizure detection 

The unpredictable nature of epileptic seizures poses significant distress and disability for both patients and caregivers. However, advancements in technology are now offering solutions. Seizure detection devices are becoming commercially available, making them more accessible for general use and enabling 24/7 monitoring outside of medical environments.

These devices can alert patients or caregivers to the early onset of seizure events. It's important to note that early detection does not equate to seizure prediction. Instead, it signifies the earliest interval when abnormal EEG activity is detected before clinical symptoms manifest.

This technology is crucial for people with epilepsy, as early detection and intervention during an epileptic attack can prevent head injuries from falls or even fatalities.

1. Key features of epilepsy seizure-detection devices:

According to current recommendations (Bruno et al., 2018, 2020; Simblett et al., 2019; Van de Vel et al., 2019), commercial epilepsy biometric devices must meet at least five key factors to ensure users' needs are addressed:

1. Reliability: This is paramount. A seizure-detection device should achieve a detection level greater than 90% and maintain a false alarm rate of less than 2 per week. These metrics are critical for ensuring the device's effectiveness and avoiding unnecessary disruptions in patients' and caregivers' lives.
2. Safety and privacy: EEG seizure-detection systems must prioritize user safety and ensure the confidentiality of recorded brain data. Robust privacy measures should be in place to protect users' sensitive information.

3. Self-management: Supporting users in self-managing seizures is essential. Devices should empower users to regulate rest and daily activities, enhancing their ability to cope with epilepsy.

4. Suitability: Design considerations, such as discreteness and low intrusiveness, are vital for ensuring device suitability. Optimal acceptability and usability are achieved when devices are comfortable and seamlessly integrate into users' lives.

5. Proof of Evidence: Marketed seizure-detection devices should be backed by scientific evidence demonstrating their reliability and efficacy. This evidence is crucial for instilling confidence in users and healthcare professionals regarding the device's capabilities.

2. Wearable seizure-detection devices currently available in the market

Ultimately, the selection of a seizure-detection device depends on various factors such as the type of seizure, individual circumstances, lifestyle, and the acceptable level of risk.

There are several wearables available in the market for early seizure detection (Bruno et al., 2020). Most of these devices utilize mobile health (mHealth) approaches, primarily focusing on recording biometric signals associated with motor manifestations of seizures (e.g., tonic-clonic seizures). However, evidence supporting the accuracy of devices for detecting seizures without major motor features is still lacking.

Below is a table listing specific instances of marketed devices for seizure detection, along with published evidence on device performance or medical device approval.

Conclusions

EEG recordings have proven invaluable in identifying abnormal neural activity associated with various types of seizures with relatively high accuracy. As EEG is safe and painless, it is often one of the first recommended tests for detecting seizures in patients diagnosed with epilepsy. A significant contribution of EEG is its ability to provide greater specificity in diagnosing types of epilepsy, thereby assisting clinicians in effectively managing the disease's treatment.

To address the overall needs of patients and their caregivers, wearable technology offers solutions for early seizure detection in real-life conditions. However, a significant challenge facing currently marketed devices is to improve seizure detection accuracy and enhance their versatility in disease self-management.

Author

Cristina Gil-López, Ph.D. (LinkedIn, Scholar, homepage) Researcher at the Psychology and Neurocognition institute (LPNC), University of Grenoble (France) / Associated Research Scientist at Polytechnic University of Valencia (Spain).

Javier Mínguez (LinkedIn, Twitter, Scholar) Associate professor of the University of Zaragoza and co-founder of Bitbrain. 

You might also be interested in

• The Use of EEG for ADHD Diagnosis and Treatment
• The Procedure and Uses of the EEG Test
• What is QEEG brain mapping and normative databases?
• EEG Synchronization With Other Biosensors (EEG, ECG, EMG, eye tracking, etc.), and Software
• How Deep Learning is changing machine learning AI in EEG data processing
• All about EEG artifacts and filtering tools
• Connectivity Layer and Other Features of EEG Headsets Explained
• EEG-based Neurotechnology for Human Enhancement and Rehabilitation
• Modern BCI-based Neurofeedback for Cognitive Enhancement
• The Ultimate Guide of Technical Features of EEG Systems
• Everything about cognitive rehabilitation in depression
• What is cognitive rehabilitation and what are the most employed techniques?
• Overview of cognitive rehabilitation and stimulation therapies in dementia

References

• Adebimpe, A., Aarabi, A., Bourel-Ponchel, E., Mahmoudzadeh, M., & Wallois, F. (2016). EEG Resting State Functional Connectivity Analysis in Children with Benign Epilepsy with Centrotemporal Spikes. Frontiers in neuroscience, 10, 143. https://doi.org/10.3389/fnins.2016.00143
• Arkilo D, Wang S, Thiele EA. Time interval required for return to baseline of the background rhythm on electroencephalogram after recorded electrographic seizures. Epilepsy Res. 2013 Sep;106(1-2):288-91. doi: 10.1016/j.eplepsyres.2013.04.007. Epub 2013 May 14. PMID: 23684125.
• A. Viloria Alebesque, A. López Bravo, E. Bellosta Diago, S. Santos Lasaosa, J.A. Mauri Llerda. Utilidad del electroencefalograma en el manejo de la epilepsia en el Servicio de Urgencias. Neurología, Volume 35, Issue 4, May 2020, Pages 238-244.https://doi.org/10.1016/j.nrl.2017.08.008
• Selim R Benbadis (2010) The tragedy of over-read EEGs and wrong diagnoses of epilepsy. Expert Review of Neurotherapeutics, 10:3, 343-346.https://doi.org/10.1586/ern.09.157
• Berg A.T., Shinnar S. Relapse following discontinuation of antiepileptic drugs: a meta-analysis. Neurology. 1994;44:601-608. 10.1212/wnl.44.4.601
• Berger, H. Über das Elektrenkephalogramm des Menschen. Archiv f. Psychiatrie 87, 527–570 (1929). https://doi.org/10.1007/BF01797193
• Blume W. T. (2010). Focal and generalized: both here and there. Epilepsy currents, 10(5), 115–117. https://doi.org/10.1111/j.1535-7511.2010.01375.x
• Claassen J, Mayer SA, Kowalski RG, Emerson RG, Hirsch LJ. Detection of electrographic seizures with continuous EEG monitoring in critically ill patients. Neurology 2004; 62: 1743– 1748 . 10.1212/01.wnl.0000125184.88797.62 
• Berg AT, Shinnar S. Relapse following discontinuation of antiepileptic drugs: a meta-analysis. Neurology.1994;44:601–8.  DOI: 10.1212/wnl.44.4.601
• Bruno, E., Viana, P. F., Sperling, M. R., & Richardson, M. P. (Accepted/In press). Seizure detection at home: Do devices on the market match the needs of people living with epilepsy and their caregivers? Epilepsia. https://doi.org/10.1111/epi.16521
• Bruno E, Simblett S, Lang A, Biondi A, Odoi C, Schulze‐Bonhage A, et al. Wearable technology in epilepsy: the views of patients, caregivers, and healthcare professionals. Epilepsy Behav. 2018;85:141–9. https://doi.org/10.1111/epi.16521
• Falco-Walter, J.J., Scheffer, I.E., and Fisher, R.S. (2018) The new definition and classification of seizures and epilepsy. Epilepsy Res. 139: 73-79.  DOI: 10.1016/j.eplepsyres.2017.11.015
• Fisher, R. S., Scharfman, H. E., & deCurtis, M. (2014). How can we identify ictal and interictal abnormal activity?. Advances in experimental medicine and biology, 813, 3–23. https://doi.org/10.1007/978-94-017-8914-1_1
• Gevins AS, Schaffer RE. A critical review of electroencephalographic (EEG) correlates of higher cortical functions. Crit Rev Bioeng. 1980;4(2):113-64. PMID: 7004774.
• Gregory RP, Oates T, Merry RT. Electroencephalogram epileptiform abnormalities in candidates for aircrew training. Electroencephalography Clin Neurophysiol. 1993 Jan;86(1):75-7. doi: 10.1016/0013-4694(93)90069-8. PMID: 7678394.
• Helmstaedter C, Witt JA. Epilepsy and cognition - A bidirectional relationship? Seizure. 2017 Jul;49:83-89. doi: 10.1016/j.seizure.2017.02.017. Epub 2017 Mar 1. PMID: 28284559
• Holmes G. L. (2013). EEG abnormalities as a biomarker for cognitive comorbidities in pharmacoresistant epilepsy. Epilepsia, 54 Suppl 2(0 2), 60–62. https://doi.org/10.1111/epi.12186
• Klein, H. (2019, August 26). Temporal Lobe Epilepsy (TLE). Epilepsy Foundation. https://www.epilepsy.com/learn/types-epilepsy-syndromes/temporal-lobe-epilepsy-aka-tle
• Iwasaki M,Pestana E,Burgess RC,Luders HO,Shamoto H,Nakasato N( 2005): Detection of epileptiform activity by human interpreters: Blinded comparison between electroencephalography and magnetoencephalography. Epilepsia 46: 59–68. https://doi.org/10.1111/j.0013-9580.2005.21104.x
• Jouny, C. C., Franaszczuk, P. J., & Bergey, G. K. (2011). Improving early seizure detection. Epilepsy & behavior : E&B, 22 Suppl 1(Suppl 1), S44–S48. https://doi.org/10.1016/j.yebeh.2011.08.029
• Manjari Tripathi and Man Mohan Mehendiratta (2014). Role of EEG in Epilepsy, Epilepsy Topics, Mark D. Holmes, IntechOpen, DOI: 10.5772/57430. Available from: https://www.intechopen.com/books/epilepsy-topics/role-of-eeg-in-epilepsy
• Pillai J, Sperling MR. Interictal EEG and the diagnosis of epilepsy. Epilepsia. 2006 ;47 Suppl 1:14-22. DOI: https://doi.org/10.1111/j.1528-1167.2006.00654.x
• Pohlmann-Eden, B., Beghi, E., Camfield, C., & Camfield, P. (2006). The first seizure and its management in adults and children. BMJ (Clinical research ed.), 332(7537), 339–342. https://doi.org/10.1136/bmj.332.7537.339
• Rogawski, M. A., Löscher, W., & Rho, J. M. (2016). Mechanisms of Action of Antiseizure Drugs and the Ketogenic Diet. Cold Spring Harbor perspectives in medicine, 6(5), a022780. https://doi.org/10.1101/cshperspect.a022780
• Shafi, M. M., Westover, M. B., Cole, A. J., Kilbride, R. D., Hoch, D. B., & Cash, S. S. (2012). Absence of early epileptiform abnormalities predicts lack of seizures on continuous EEG. Neurology, 79(17), 1796–1801. https://doi.org/10.1212/WNL.0b013e3182703fbc
• Sharbrough FW. Scalp recorded ictal patterns in focal epilepsy. J Clin Neurophysiol 1993; 10:262-267. DOI: 10.1097/00004691-199307000-00002
• Sim I. Mobile Devices and Health. N Engl J Med. 2019 Sep 5;381(10):956-968. doi: 10.1056/NEJMra1806949. PMID: 31483966.
• Simblett SK, Bruno E, Siddi S, Matcham F, Giuliano L, López JH, et al. Patient perspectives on the acceptability of mHealth technology for remote measurement and management of epilepsy: a qualitative analysis. Epilepsy Behav. 2019;97:123–9. DOI: 10.1016/j.yebeh.2019.05.035
• Smith, S J M. “EEG in the diagnosis, classification, and management of patients with epilepsy.” Journal of neurology, neurosurgery, and psychiatry vol. 76 Suppl 2,Suppl 2 (2005): ii2-7. doi:10.1136/jnnp.2005.069245
• Sundaram M, Sadler RM, Young GB, Pillay N. EEG in Epilepsy: Current Perspectives. Can. J. Neurol. Sci. 1999; 26: 255-262
• National Institute for Clinical Evidence. NICE guidelines: the diagnosis and management of the epilepsies in adults and children in primary and secondary care. London: NICE, 2004
• Van de Vel A, Smets K, Wouters K, Ceulemans B. Automated non‐EEG based seizure detection: do users have a say? Epilepsy Behav. 2016;62:121–8. DOI: 10.1016/j.yebeh.2016.06.029