Long-Term Monitoring & Ambulatory EEG
Prolonged EEG monitoring — whether performed in a specialized video-EEG monitoring (VEM) unit or in the ambulatory setting — greatly expands the diagnostic capabilities of EEG beyond what routine recordings can achieve. Video-EEG monitoring is the gold standard for presurgical localization of seizure onset, differentiation of epileptic seizures from psychogenic nonepileptic events (PNES), and precise classification of seizure types. Ambulatory EEG provides an accessible alternative for patients who do not require medication tapering or intensive observation, and it offers the advantage of recording in the patient's natural environment. Emerging wearable EEG technologies, subcutaneous implantable devices, and automated seizure detection algorithms are reshaping the landscape of long-term epilepsy monitoring and may enable truly continuous, real-world seizure tracking in the future.
Bottom Line
- Video-EEG monitoring (VEM) is essential for presurgical evaluation, seizure classification, and diagnosis of PNES — estimated to account for at least 20% of patients referred for drug-resistant seizures
- ASM reduction during VEM accelerates seizure capture but requires careful safety protocols for seizure clusters, secondary generalization, and SUDEP risk
- Ambulatory EEG (1–4 days) has significantly higher sensitivity (72% in 1 day) than routine EEG (11% on first study) for detecting epileptiform abnormalities after a first unprovoked seizure
- Ambulatory EEG limitations: No on-site clinical testing during events; artifact burden from daily activities; limited ability to reduce medications safely
- Subcutaneous EEG devices are in development for ultra-long-term monitoring (months to years) and may enable seizure forecasting, early intervention, and personalized treatment optimization
- Event detection algorithms powered by artificial intelligence are approaching expert-level performance for routine EEG interpretation and will likely expand access to expert-level analysis
Video-EEG Monitoring Unit
Purpose and Setup
The video-EEG monitoring unit is a specialized inpatient facility designed for simultaneous, continuous recording of EEG and time-locked video of the patient's behavior. This combination allows precise correlation of electrographic events with clinical manifestations.
- Electrode system: Standard 10–20 system with additional electrodes as needed (anterior temporal T1/T2, sphenoidal); electrodes applied with collodion for long-term adhesion
- Video: Continuous infrared-capable video recording with audio; camera positioned to capture the patient's full body and face
- Nursing staff: Specialized nurses trained to administer clinical testing during seizures (orientation, language, motor function) to assess awareness and lateralizing signs
- Safety measures: Padded bed rails, seizure response protocols, oxygen at bedside, suction equipment, emergency medication (benzodiazepines) readily available; continuous monitoring by technologists or remote staff
- Duration: Typically 3–7 days; may be extended if insufficient events are captured or if additional data is needed for surgical planning
Indications
| Indication | Goal | Expected Outcome |
|---|---|---|
| Presurgical evaluation | Localize seizure onset zone; establish concordance with imaging and neuropsychological data | Ictal recordings with clear onset region; correlation with MRI lesion, PET/SPECT findings; determination of surgical candidacy |
| Seizure classification | Distinguish focal vs. generalized; classify specific seizure types (e.g., absence vs. focal impaired awareness) | Accurate syndrome diagnosis guiding optimal ASM selection; identification of multiple seizure types in the same patient |
| PNES diagnosis | Capture typical events and demonstrate absence of ictal EEG correlate | Definitive diagnosis when a typical event is captured without electrographic seizure activity; enables appropriate psychiatric/psychological referral |
| Quantifying seizure burden | Determine true frequency of seizures, including subclinical events | More accurate seizure frequency data for treatment decisions; identification of unrecognized nocturnal or subclinical seizures |
| Treatment response assessment | Evaluate response to medication changes or neurostimulation | Objective documentation of seizure frequency and interictal burden before and after intervention |
PNES: Critical Diagnostic Considerations
- Psychogenic nonepileptic events account for at least 20% of patients with suspected intractable seizures referred for epilepsy surgery evaluation
- A seizure must be confirmed as epileptic before epilepsy surgery is considered — VEM is the definitive test for distinguishing PNES from epileptic seizures
- PNES and epilepsy can coexist in the same patient (estimated in 5–20% of PNES patients); both types of events must be captured to establish the diagnosis
- Features suggestive of PNES on VEM: eyes closed during event, prolonged duration (>2 minutes for convulsive events), waxing-and-waning intensity, asynchronous bilateral limb movements, pelvic thrusting, preserved awareness despite bilateral motor activity
- Caution: Some epileptic seizures (especially frontal lobe) can mimic PNES with bizarre motor manifestations and no clear ictal EEG correlate; absence of surface EEG changes does not definitively exclude an epileptic seizure
ASM Reduction Protocols During Monitoring
Reduction or withdrawal of antiseizure medications (ASMs) during VEM is a common strategy to provoke seizures and increase the yield of monitoring. This must be balanced against the risks of seizure clusters, status epilepticus, and injury.
Approaches to ASM Reduction
- Gradual taper: ASMs are reduced stepwise over 2–4 days; the most common approach; allows time to observe the effect of each reduction
- Partial reduction: Reducing to 50% of home dose rather than complete withdrawal; appropriate for patients at high risk for generalized tonic-clonic seizures or status epilepticus
- Complete withdrawal: All ASMs discontinued; higher yield but increased risk; generally reserved for selected patients with a clear safety plan
- Selective withdrawal: Reducing only one agent in polytherapy while maintaining others; useful when specific seizure types need to be provoked
Safety Considerations
| Risk | Mitigation Strategy |
|---|---|
| Seizure clusters | Pre-specified rescue medication protocol (IV lorazepam, rectal diazepam); seizure count thresholds for reloading ASMs; continuous nurse observation |
| Status epilepticus | Emergency SE protocol readily available; IV access maintained; airway management equipment at bedside |
| Bilateral tonic-clonic seizures | Limit the number of GTC seizures allowed (typically 1–3); reload ASMs after threshold is reached; ensure postictal monitoring for respiratory compromise |
| SUDEP risk | Continuous pulse oximetry and cardiac monitoring; positioning protocols (avoid prone); immediate response to prolonged GTC seizures or desaturation; nighttime supervision |
| Falls and injury | Padded bed rails; helmet if ambulatory monitoring; call button accessible; restricted ambulation during high-risk periods |
Clinical Pearl: Value of Longer Recording Duration
- VEM not only captures ictal events but also reveals the full spectrum of interictal abnormalities — generalized spike-and-wave, polyspike-and-wave, generalized paroxysmal fast activity, focal slowing, and focal spikes may all coexist in the same patient
- The longer recording duration allows for quantification of electrographic abnormalities and assessment of medication response
- Multiple seizure types may be identified that were not reported by the patient, as many seizures (especially brief absences and nocturnal seizures) go unrecognized
- Documentation of seizure semiology on video is invaluable for surgical planning and may reveal lateralizing signs (forced head version, unilateral automatisms, postictal Todd paralysis) that the patient does not report
Ambulatory EEG
Ambulatory EEG (aEEG) allows prolonged recording in the outpatient setting, typically over 1–4 days, in the patient's home environment. This modality has become increasingly practical with improvements in portable recording equipment.
Indications
- Patients who do not require medication tapering or inpatient-level safety monitoring
- Events provoked by routine daily activities, stress, or specific environmental triggers not reproducible in the hospital
- Patients who cannot tolerate an inpatient stay (children, elderly, patients with significant comorbidities or anxiety)
- Assessment of interictal background for unrecognized or subclinical seizures
- Evaluation when routine outpatient EEG has been nondiagnostic
Yield and Performance
| Study Type | Duration | Sensitivity for IEDs | Comparison |
|---|---|---|---|
| First routine EEG | 20–40 min | 11% | Same population of patients with first unprovoked seizure (Hernandez-Ronquillo et al. 2023) |
| Second routine EEG | 20–40 min | 22% | |
| 1-day ambulatory EEG | ~24 hours | 72% | |
| 4-day ambulatory EEG | ~96 hours | 80–90% (estimated) | Extended ambulatory recording further increases yield |
Limitations
- No on-site clinical testing: Events cannot be assessed in real time by trained staff; clinical details rely on patient/family pushbutton event markers and diary entries
- Video quality: Many ambulatory systems include a small portable camera, but video quality and coverage are significantly inferior to inpatient VEM
- Artifact: Recording in the home environment produces more artifact from daily activities (eating, walking, driving); electrode integrity may degrade over multiple days
- Safety: No immediate medical intervention available for prolonged seizures or status epilepticus; ASM reduction is generally not performed during ambulatory monitoring
- Technical support: Electrode repair or troubleshooting requires either a return visit or a home technologist visit
Wearable EEG Devices and Emerging Technology
The future of epilepsy monitoring is moving toward continuous, real-world seizure detection using miniaturized, wearable, or implantable devices. Several technologies are in various stages of development and clinical validation.
Subcutaneous EEG
- Surgically implanted subcutaneous EEG devices have been proposed for ultra-long-term monitoring (months to years)
- Several devices are in various stages of development in the United States
- Potential applications include: characterization of infrequent events that are difficult to capture during ambulatory or inpatient studies; long-term quantification of seizure burden for treatment optimization; identification of patient-specific seizure patterns for quality-of-life interventions; caregiver alerting when a seizure occurs; and seizure forecasting to warn patients before a seizure occurs
- Early studies have demonstrated the feasibility of individualized seizure forecasting models using subcutaneous EEG data
Behind-the-Ear and Wearable Devices
- Miniaturized EEG devices worn behind the ear or embedded in headbands are being developed for continuous ambulatory monitoring
- These devices typically record from 1–4 channels and use machine learning algorithms for automated seizure detection
- Limitations include limited spatial coverage (may miss focal seizures distant from recording electrodes) and susceptibility to movement artifact
Artificial Intelligence in EEG Interpretation
- Data exist regarding the ability of artificial intelligence (AI) to approach expert levels of interpretation for routine EEG data
- AI-based automated interpretation of clinical EEGs showed agreement with expert neurophysiologists comparable to interrater agreement among experts themselves
- As these technologies advance, they will likely enable expert-level interpretation of large quantities of data and expand accessibility to expert-level EEG interpretation in underserved settings
- Current automated seizure detection algorithms have sensitivities of 75–90% but generate false-positive alerts that require human verification
Choosing Between Inpatient VEM and Ambulatory EEG
- Choose inpatient VEM when: Presurgical evaluation is needed; PNES diagnosis is suspected; medication tapering is required to provoke events; seizure clusters or status epilepticus risk is significant; precise ictal semiology documentation is required
- Choose ambulatory EEG when: Interictal characterization is the primary goal; events are provoked by daily activities or environmental triggers; the patient cannot tolerate inpatient admission; routine EEG has been nondiagnostic and extended recording is needed; the clinical question is about the burden of subclinical seizures in a known epilepsy patient
- Consider ambulatory EEG as a bridge: In patients who need further evaluation but face long wait times for inpatient VEM, ambulatory EEG can provide useful interictal data and may capture events that clarify the diagnosis
Event Detection and Seizure Diaries
Accurate tracking of seizure events during prolonged monitoring is essential for both clinical decision-making and research.
- Pushbutton events: Patients and families are trained to press an event button on the EEG recorder when a seizure or suspicious event occurs; this creates a time-stamp that the EEG reviewer can immediately examine
- Seizure diaries: Written or electronic logs maintained by the patient and/or family; important for events that occur when the pushbutton is not accessible
- Automated detection: Modern EEG recording systems include built-in seizure detection algorithms that flag potential seizures for review; these algorithms reduce the time required to review prolonged recordings but have variable sensitivity and specificity depending on the seizure type and patient
- Discrepancy between self-reporting and EEG: Studies consistently show that patients undercount their seizures, particularly nocturnal events and brief seizures without prominent motor manifestations; cEEG or VEM often reveals significantly more events than reported
References
- Hernandez-Ronquillo L, Thorpe L, Feng C, et al. Diagnostic accuracy of ambulatory EEG vs routine EEG in patients with first single unprovoked seizure. Neurol Clin Pract. 2023;13(3):e200160.
- Krumholz A. Nonepileptic seizures: diagnosis and management. Neurology. 1999;53:s76–s78.
- Fisher RS, Acevedo C, Arzimanoglou A, et al. ILAE official report: a practical clinical definition of epilepsy. Epilepsia. 2014;55(4):475–482.
- Peltola ME, Leitinger M, Halford JJ, et al. Routine and sleep EEG: minimum recording standards of the IFCN and ILAE. Epilepsia. 2023;64(3):602–618.
- Duun-Henriksen J, Baud M, Richardson MP, et al. A new era in electroencephalographic monitoring? Subscalp devices for ultra-long-term recordings. Epilepsia. 2020;61(9):1805–1817.
- Viana PF, Pal Attia T, Nasseri M, et al. Seizure forecasting using minimally invasive, ultra-long-term subcutaneous electroencephalography: individualized intrapatient models. Epilepsia. 2022.
- Tveit J, Aurlien H, Plis S, et al. Automated interpretation of clinical electroencephalograms using artificial intelligence. JAMA Neurol. 2023;80(8):805.
- Burkholder DB, Britton JW, Rajasekaran V, et al. Routine vs extended outpatient EEG for the detection of interictal epileptiform discharges. Neurology. 2016;86(16):1524–1530.
- Losey TE, Uber-Zak L. Time to first interictal epileptiform discharge in extended recording EEGs. J Clin Neurophysiol. 2008;25(6):357–360.
- Gloss D, Pargeon K, Pack A, et al. Antiseizure medication withdrawal in seizure-free patients: practice advisory update summary. Neurology. 2021;97(23):1072–1081.
- Terman SW, van Griethuysen R, Rheaume CE, et al. Antiseizure medication withdrawal risk estimation and recommendations. Epilepsia Open. 2023;8(2):386–398.
- Ebersole JS, Pacia SV. Localization of temporal lobe foci by ictal EEG patterns. Epilepsia. 1996;37(4):386–399.
- Laskowitz DT, Sperling MR, French JA, O'Connor MJ. The syndrome of frontal lobe epilepsy: characteristics and surgical management. Neurology. 1995;45(4):780–787.
- Quesney LF. Preoperative electroencephalographic investigation in frontal lobe epilepsy. Can J Neurol Sci. 1991;18(S4):559–563.
- Seneviratne U, Hepworth G, Cook M, D'Souza W. Atypical EEG abnormalities in genetic generalized epilepsies. Clin Neurophysiol. 2016;127(1):214–220.