
A wearable electroencephalograph (EEG) device that leverages artificial intelligence allows for precise, 12-to-24-h home monitoring of absence seizures, enabling rapid and tailored adjustments to antiseizure medication (ASM) and eliminating the need for further hospital admissions, as shown in an exploratory study.
The device consists of four electrodes placed on the mastoid bone behind the ears, two on each side of the head, approximately 5 cm apart. These electrodes are connected to the mobile EEG device, which is attached to the upper back via an adhesive. The device records two-channel EEG data at 250 Hz for up to 30 hours. Additionally, it incorporates an inertial measurement unit that uses accelerometer and gyroscope sensors to measure patient movement. A docking station serves as charging unit for the device’s battery and a hub for automatic EEG data upload to the cloud via Wi-Fi.
Patients are instructed to use the device between 12 and 24 h within 1 week after adjustments in ASM, ideally on days when they could stay at home. After each use, the device should be returned to its docking station to upload data for the clinical team’s review and analysis. An algorithm using EEG and movement data from accelerometer and gyroscope filters out motion artifacts and reduces the amount of data needed to be reviewed.
The aim is to optimize ASM and obtain seizure freedom for the patient, the investigators said. “In clinical practice, physicians often rely on seizure diaries and patient reports to determine treatment responses. However, between 74 percent and 94 percent of absence seizures are not reported accurately… [Furthermore,] 24-h video-EEG monitoring in the hospital is not feasible everywhere, and where possible, it is costly, with prolonged waiting times.”
The wearable EEG device was designed to help to overcome those limitations, they pointed out. If the wearable EEG device detects ongoing seizures, this immediately triggers further treatment adjustments by the clinician, they added. This cycle of treatment adjustment followed by device-based monitoring continues until seizure freedom is achieved.
Proof of concept
In the current exploratory case series, the investigators evaluated the utility of the wearable EEG device in terms of improving absence seizure detection at home and facilitating timely ASM adjustments during follow-up. They also assessed the performance of the algorithm against that of a neurologist with EEG training who initially reviewed the signals. Absence seizures were defined as 3-Hz (2.5–5.5-Hz) spike-and-wave discharges (SWDs) lasting 3 s or longer.
Nineteen patients (median age 24 years, 63.2 percent female, median age at epilepsy onset 13 years) with idiopathic generalized epilepsy and typical absence seizures were included in the analysis. Of these patients, 13 (68 percent) had juvenile absence epilepsy and six (32 percent) had juvenile myoclonic epilepsy. Eleven patients (58 percent) had refractory epilepsy. The median number of ASMs at baseline was 2. Two patients were not on ASM (one with a recently diagnosed epilepsy, and another had discontinued ASM several years previously), and one patient had a vagus nerve stimulator. [Epilepsia 2025;doi:10.1111/epi.18384]
Over a median follow-up of median of 5 months, 15 patients (79 percent) achieved freedom from seizure, including seven of 11 (63 percent) who had refractory epilepsy. The median recording time for each session of device use was 21.3 h. Ten patients relapsed after a median of 1–2 recordings that showed no 3-Hz SWDs.
Four patients (21 percent) had patch imprinting and skin redness after use of the device, although these side effects did not prevent continued use of the device. Meanwhile, four patients (21 percent) experienced side effects related to increases in ASM. Of these, one required a decrease in the ASM dose and another discontinued the medication.
With regard to the EEG data and algorithm performance, manual review of the EEG data led to the identification of 806 3-Hz SWDs of ≥3 s. The algorithm substantially improved the review process for 24-hour wEEG data, reducing the median review time from 27 to 4.3 min. The corresponding values for sensitivity, precision, F1-score, and false positives per hour were 0.8, 0.95, 0.87, and 0.007, respectively.
Recommended use cases
Taken together, the data demonstrate the utility of the wearable EEG device and the review algorithm in at-home absence seizure detection, according to the investigators. However, they noted that despite the unobtrusive design of the device, there were periods when patients were unwilling to use it, specifically during holidays.
The investigators recommended the use of the device for patients with typical absence seizures under certain circumstances.
“First, in recently diagnosed patients, the device allows defining the lowest ASM dose required to reach seizure freedom. Second, in seizure-free patients with side effects, the device can be used to assess the impact of switching to another ASM, favouring the objective assessment of seizure freedom afterward,” the investigators said.
“Third, in patients with refractory absence seizures who have tried several ASMs, it is often impossible to define the best combination and doses. In our experience with [one patient], the device facilitated testing the effect of ASM tapering due to a suspected lack of efficacy and weight gain, only to find that the patient required the medication in lower doses,” they pointed out, noting that this was a patient who was not able to achieve seizure freedom with standard clinical practices over the preceding 4 years.
The investigators believed that the use of the wearable EEG device may have important economic implications, helping reduce costs and waiting times in epilepsy monitoring units.