Enhanced EEG Signal Processing for Accurate Epileptic Seizure Detection

Document Type

Article

Publication Title

SN Computer Science

Abstract

Electroencephalography (EEG) is a fundamental technique for epileptic seizure detection due to its ability to capture brain activity. Symmetry and asymmetry in EEG signals serve as indicators of seizure events, as normal EEGs exhibit bilateral patterns that may become asymmetric during seizures. The paper discusses applying symmetry/asymmetry analysis in the diagnostic process for epilepsy. To improve how accurate detection can be, Improved Feature Space Method (ICFS) with DWT extracts features from the datasets in time, frequency and entropy domains. The features are processed by various Support Vector Machine (SVM) models during ensemble training. Tests on standard EEG datasets found that our method achieved 97% accuracy during validation and outperformed both traditional correlation methods and advanced seizure detection approaches. The study reveals the model works effectively and can accurately detect seizures.

DOI

10.1007/s42979-025-04148-1

Publication Date

8-1-2025

This document is currently not available here.

Share

COinS