A quadratic linear-parabolic model-based classification to detect epileptic EEG seizures

TL;DR

This paper presents a model-based EEG seizure detection method that filters signals, fits a quadratic model, and classifies events using random forests, achieving high accuracy and speed.

Contribution

It introduces a novel EEG classification approach combining filtering, quadratic model fitting, and statistical features with machine learning for seizure detection.

Findings

92% sensitivity in seizure detection

96% specificity achieved

94.1% overall accuracy

Abstract

The two-point central difference is a common algorithm in biological signal processing and is particularly useful in analyzing physiological signals. In this paper, we develop a model-based classification method to detect epileptic seizures that relies on this algorithm to filter EEG signals. The underlying idea is to design an EEG filter that enhances the waveform of epileptic signals. The filtered signal is fitted to a quadratic linear-parabolic model using the curve fitting technique. The model fitting is assessed using four statistical parameters, which are used as classification features with a random forest algorithm to discriminate seizure and non-seizure events. The proposed method was applied to 66 epochs from the Children Hospital Boston database. Results show that the method achieves fast and accurate detection of epileptic seizures, with a 92% sensitivity, 96% specificity, and 94.1% accuracy.