Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
May 6, 2026Open Access

Simulated reduced-channel EEG montages yield modest performance drops in automated seizure detection versus full montages.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Implantable sub-scalp EEG systems show promise for long-term seizure monitoring in epilepsy, but how automated seizure detection performance varies by montage configuration is unknown.

Does automated seizure detection performance differ between full and reduced montages in patients with epilepsy?

Population

436 patients (466 admissions) with at least one annotated seizure and one interictal clip

Comparison

Full vs reduced montages simulating published sub-scalp devices

Design

Single-center retrospective cross-sectional simulation study

Key result

Automated seizure detection using simulated reduced-channel montages showed only modest performance decreases (absolute F1 change -0.09 to 0.014) compared to full montages.

Authors

JKJoe KojimaHSHaoer ShiSJSvanik Jaikumar

Discussion

Loading...

Member takes

Overview

May support sub-scalp device feasibility for long-term monitoring; leaves open prospective validation before clinical use.

Study Design

Type

Cross-Sectional (n=436)

Multicenter

No

Structured PICO

Does automated seizure detection performance differ between full and reduced montages in patients with epilepsy?

P
Population
436 patients (466 admissions) with epilepsy, mean age 39.0 years, 54.4% female, with at least one annotated seizure and one interictal clip ≥20 minutes from any seizure.
I
Intervention
Computational simulation of published sub-scalp device montages using standard 10-20 EEG channels
C
Comparator
Full montage EEG
O
Outcome
Event-based F1 scores evaluated for three published seizure detectors (SVM, SPaRCNet, NDD) across montagessurrogate

Main Result

Effect estimate: absolute F1 change -0.09 to 0.014

Automated seizure detection algorithms perform with comparable accuracy on reduced-channel montages simulating sub-scalp devices, supporting their feasibility for ultra-long-term monitoring.

Cite This Study

Kojima et al. (2026) conducted a cross-sectional in Epilepsy (n=436). Reduced montages simulating published sub-scalp devices vs. Full montage (standard 10-20 EEG channels) was evaluated on Event-based F1 scores evaluated for three published seizure detectors (absolute F1 change -0.09 to 0.014). Automated seizure detection using simulated reduced-channel montages showed only modest performance decreases (absolute F1 change -0.09 to 0.014) compared to full montages.

synapsesocial.com/papers/6a0ee778b9cfc04f9247ae7ehttps://doi.org/10.64898/2026.05.05.26352477
View Full Paper
Ask AI
Bookmark
Share