An observational study in Epilepsy, Absence Epilepsy and Juvenile Absence Epilepsy, sponsored by Universitaire Ziekenhuizen KU Leuven. Not yet recruiting. Open to participants aged 6 Years and older. Per ClinicalTrials.gov, last updated 2026-11-18.
Sponsored by Universitaire Ziekenhuizen KU Leuven · Observational
Help build ClinicalTrials.gg — short surveys, remote interviews or testing, in your language. Take part →
Typical absence seizures are the hallmark of idiopathic generalized epilepsy syndromes (CAE, JAE, JME) but are frequently underreported in patient diaries. Clinical studies have shown that 74% to 94% of absence seizures go unreported by patients or caregivers. While 24-hour video-EEG (vEEG) is the gold standard for quantification, it is expensive, hospital-based, and lacks the temporal resolution required for long-term management.
Previous studies successfully validated a wearable EEG (wEEG) solution using a behind-the-ear (BTE) form factor (Byteflies Sensor Dot). In the pilot study, the authors demonstrated that a reduced electrode configuration combined with ML algorithms could detect absence seizures with 98% sensitivity in the hospital setting and enable effective home-based treatment monitoring.
To expand the applicability of this method, validation of additional form factors is required. This study aims to extend our validation framework to a novel wEEG device that utilizes an ear configuration (earEEG).
Background, rationale and risk assessment
1.1 Disease Background Typical absence seizures are the clinical hallmark of idiopathic generalized epilepsy syndromes, specifically Childhood Absence Epilepsy (CAE), Juvenile Absence Epilepsy (JAE), and Juvenile Myoclonic Epilepsy (JME)(1). These syndromes account for up to 20% of all people living with epilepsy, primarily affecting children, adolescents, and young adults(2). Clinically, absences are characterized by a transient loss of consciousness accompanied by generalized 3 Hz (2.5-5.5 Hz) spike-and-wave discharges (SWD) on the electroencephalogram (EEG). These seizures can have significant cognitive, affective, and behavioral comorbidities, particularly when not adequately controlled(1,3-5).
1.2 Limitations of the Current Standard of Care Accurate quantification of the seizure burden is essential for optimizing the treatment with anti-seizure medications (ASM). The current standard of care relies on seizure diaries. However, previous studies indicate that the sensitivity of self-reporting for absence seizures ranges from 6% to 26% compared with video-EEG(6,7). i.e., seizure diaries have low reliability.
Therefore, patients are intermittently followed with a 24-hour video electroencephalogram (vEEG), which has, among other limitations, high costs (approximately €1,200 per session in Belgium), long waiting times, and limited ecological validity, as they provide only a snapshot of the patient's condition that may not capture cyclical seizure variability(8).
This study seeks to validate an objective, automated counting method using a wearable EEG (wEEG) device, first in a hospital setting and then at home.
1.3 Rationale for Wearable EEG wEEG devices offer a promising alternative to full-scalp vEEG for long-term monitoring. By using a limited number of electrodes combined with Machine Learning (ML) algorithms, wEEG can enable intermittent or continuous, objective seizure counting outside the hospital.
Researchers from ESAT-STADIUS \& UZ Leuven have previously validated this approach using a behind-the-ear (BTE) wEEG device (The Sensor Dot, Byteflies, Antwerpen, Belgium). In a pilot study, an ML algorithm developed by the group achieved 98% sensitivity for detecting absence seizures in a hospital setting compared to standard EEG(6). Furthermore, a subsequent home-based pilot demonstrated the feasibility of using this system for treatment optimization(9).
1.4 Justification for the Current Investigation While the BTE form factor has been validated, it may not be the optimal solution for all patients due to cosmetic concerns, comfort, or lifestyle constraints. To facilitate widespread adoption and personalized care, it is necessary to validate alternative wEEG form factors, such as earEEG.
The device under investigation in this study, the NeuroBuds™, uses a novel ear electrode configuration that lacks rigorous clinical validation against the gold standard for absence seizure detection.
Therefore, the primary purpose of this study is to validate the signal quality and diagnostic accuracy of the ESAT-STADIUS ML algorithms on the signals recorded with the Neurobuds against simultaneous standard vEEG. This validation is a prerequisite for developing a device-agnostic ML algorithm capable of processing data from various manufacturers and electrode montages.
1.5 Benefit/Risk Assessment The IMD is CE-marked, non-invasive, and poses minimal risk to participants, comparable to that of standard scalp EEG electrodes. The potential risks are limited to mild skin irritation or discomfort at the electrode sites. The potential benefit is the validation of a less invasive monitoring alternative. Successful validation may allow for future home-based seizure quantification.
2.1 Investigation objectives
Primary Objective The primary objective of this investigation is to validate the diagnostic accuracy of an investigational ear wEEG device, namely the NeuroBuds (Cebreo Medical), and an ML algorithm in detecting typical absence seizures. Specifically, the study aims to determine the sensitivity, precision, F1-score, and the FAR/24h of the ML seizure detection algorithm applied to signals recorded by the IMD, compared with simultaneous gold-standard video-EEG (vEEG) annotated by a neurologist.
Secondary Objectives
Hypothesis In the hospital setting, the ML algorithm will achieve seizure-detection sensitivity of ≥90% (relative to simultaneous neurologist-annotated vEEG) after analyzing the data obtained with the IMD. These results should be comparable to those previously achieved with the BTE wEEG.
2.2 Primary endpoint
The primary endpoint is the diagnostic performance of the automated absence seizure detection algorithm using the investigational device data, measured by:
2.3 Secondary endpoints
Signal Quality Metrics: Comparative hourly analysis of root mean square (RMS) and signal-to-noise ratio in the alpha band (SNR-Alpha) in the signals originated with the IMD channels and the spatially corresponding channels of the standard 10-20 EEG.
Usability Scores: Quantitative assessment of user experience (patients/caregivers) using standardized questionnaires (i.e., System Usability Scale, final questionnaire-comfort \& user experience, and User Perspective Questionnaire).
Data Yield: The percentage of interpretable (artifact-free) data obtained during unsupervised home recording sessions.
References
25 studies on the registry are indexed under Epilepsy, Absence; 2 are open to participants now.
Browse Epilepsy, Absence studies →Universitaire Ziekenhuizen KU Leuven is the lead sponsor of 965 studies on the registry; 269 are open to participants now.
Of its 5 completed or terminated interventional studies of FDA-regulated products, 0 (0%) have results posted.
Compared within Epilepsy, Absence, this study's primary condition as ClinicalTrials.gg classifies it; the registry's first condition term is Epilepsy. The counts are of registry records indexed under each term, refreshed daily. How we classify →
Adults and children with idiopathic generalized epilepsy syndromes and typical absence seizures.
Go through the criteria one by one — yes, no or not sure — and take the list to your doctor or the study team. Kept in this browser only.
Exclusion Criteria:
Underlined terms in the criteria open these too. Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Before you join, the study team explains what taking part involves — its purpose, what happens, the risks and possible benefits — and answers your questions, so you can decide freely. Consent is usually recorded on a form (the informed consent form, ICF) signed by you or, where the law allows, by someone legally authorized to decide for you.
You can keep asking questions during the study, you are told of new information that matters, and you can choose to leave at any time.
Ways of preventing pregnancy. “Highly effective” methods fail less than 1% of the time a year when used correctly — for example an IUD or an implant.
Criteria usually say which methods count and how long to keep using them after the last dose, sometimes for male participants too. Some studies accept a partner’s vasectomy if it was confirmed successful and that person is the participant’s only sexual partner.
A drug or device being studied in a clinical trial.
It may be new, or already approved and being studied for another use, dose or form. Criteria often leave out people who took part in another study of one within a recent period, commonly 30 days or five half-lives.
Able to become pregnant: from the first period until menopause, unless permanently unable to through surgery (removal of the womb, or of both ovaries or both fallopian tubes).
Criteria for people of childbearing potential usually ask for a negative pregnancy test and for contraception during the study and for a time after it.
Informed consent
Before you join, the study team explains what taking part involves — its purpose, what happens, the risks and possible benefits — and answers your questions, so you can decide freely. Consent is usually recorded on a form (the informed consent form, ICF) signed by you or, where the law allows, by someone legally authorized to decide for you.
You can keep asking questions during the study, you are told of new information that matters, and you can choose to leave at any time.
Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Contraception
Ways of preventing pregnancy. “Highly effective” methods fail less than 1% of the time a year when used correctly — for example an IUD or an implant.
Criteria usually say which methods count and how long to keep using them after the last dose, sometimes for male participants too. Some studies accept a partner’s vasectomy if it was confirmed successful and that person is the participant’s only sexual partner.
Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Investigational drug or device
A drug or device being studied in a clinical trial.
It may be new, or already approved and being studied for another use, dose or form. Criteria often leave out people who took part in another study of one within a recent period, commonly 30 days or five half-lives.
Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Childbearing potential
Able to become pregnant: from the first period until menopause, unless permanently unable to through surgery (removal of the womb, or of both ovaries or both fallopian tubes).
Criteria for people of childbearing potential usually ask for a negative pregnancy test and for contraception during the study and for a time after it.
Explained by clinicaltrials.gg, not part of the registry record. The study team decides who can take part.
Adults and children with idiopathic generalized epilepsy syndromes, i.e., Childhood Absence Epilepsy, Juvenile Absence Epilepsy, and Juvenile Myoclonic Epilepsy, with active disease and having typical absences. Prospective, single-center, open-label, comparative diagnostic accuracy study. Study participants will be contacted during the screening phase based on the scheduled admission for 24-hour vEEG in the epilepsy monitoring units of UZ Leuven. After screening and signing the IFC, they will be asked to come to the hospital for ear-mold imprinting. These prints will serve as the basis for customizing the headset. The study will employ a hybrid design, consisting of two phases: Phase A (In-Hospital Validation: participants will wear the IMD simultaneously with the standard 10-20 scalp EEG montage. This provides the ground truth labels for validation. Phase B (At home): participants will wear the device at least once a week for 3 weeks to assess real-world usability
Sensitivity (Recall)
The proportion of true absence seizures (defined as 3Hz Spike and Wave Discharges ≥ 3 seconds identified on the gold-standard vEEG) that are correctly detected by the wEEG algorithm.
Time frame: 1 day, during simultaneous use of the IMD and 25-channel scalp EEG (in-hospital phase)
Precision
The proportion of positive predictions that were actually correct when compared to the vEEG.
Time frame: 1 day, during simultaneous use of the IMD and 25-channel scalp EEG (in-hospital phase)
F1-score
The harmonic mean metric used to evaluate the overall performance of the ML model. It uses both precision and sensitivity.
Time frame: 1 day, during simultaneous use of the IMD and 25-channel scalp EEG (in-hospital phase)
False Alarm Rate in 24 hours
The number of false positive detections per 24 hours of recording.
Time frame: 1 day, during simultaneous use of the IMD and 25-channel scalp EEG (in-hospital phase)
Signal quality
Comparative hourly analysis of root mean square (RMS) and signal-to-noise ratio in the alpha band (SNR-Alpha) in the signals originated with the IMD channels and the spatially corresponding channels of the standard 10-20 EEG.
Time frame: 1 day, during simultaneous use of the IMD and 25-channel scalp EEG (in-hospital phase)
Usability Scores
Quantitative assessment of user experience (patients/caregivers) using standardized questionnaires (i.e., System Usability Scale, final questionnaire-comfort \& user experience, and User Perspective Questionnaire).
Time frame: 3 weeks, when the IMD will be used during 24 hours continuously at least once a week.
Data Yield
The percentage of interpretable (artifact-free) data obtained during unsupervised home recording sessions.
Time frame: 3 weeks, when the IMD will be used during 24 hours continuously at least once a week.
No study locations are listed for this record.
Plan to share: Yes — Two years after the analysis, we will post the database, along with all measurements and code, on OpenNeuro.
Supporting information: Analytic code
Time frame: Two years after the analysis, we will post the database, along with all measurements and code, on OpenNeuro.
Access criteria: It will be open access
No publications or documents are linked to this record.
Start is estimated for Jan 2027; primary completion for Dec 2027, 11 months later.
How these figures are worked out, and which studies are left out of them.
From the registry record's own update history. This site started tracking changes on Sep 25, 2026; for anything earlier, see the record history on ClinicalTrials.gov ↗
This study is not yet recruiting, as verified in Nov 2026. You cannot join it, but the record below documents what was studied.
Get an email when the registry record changes — status, dates, results — or when someone posts here.
Sign in to followQuestions and observations about this study, from anyone following it. Not medical advice, and not a channel to the study team — their contact details are on the registry record.
Sign in to join the discussion. Reading takes no account; posting does. You choose a display name, and a pseudonym is the default.
Nothing here yet. If you are running this trial, taking part in it, or weighing whether to, this is the place to say so.
Universitaire Ziekenhuizen KU Leuven →