Classification of EEG signals using fractals Dimensions for diagnosing epilepsy

Authors

  • Department of Computer Science, College of Computer Science and Mathematics, University of Thi- Qar, Iraq
  • Department of Computer Science, College of Computer Science and Mathematics, University of Thi- Qar, Iraq

DOI:

https://doi.org/10.32792/jeps.v13i1.244

Keywords:

Electroencephalograms (EEG), epilepsy, temporal frequency image (TFI), fractal dimension (FD), short-time Fourier transform (STFT), K-nearest neighbor(KNN), Support vector machine (SVM), and tree decision(TD).

Abstract

Electroencephalography signals derived from electrical activity in the brain are commonly used to
diagnose neurological diseases. These signals reveal electrical activity in the brain and provide
information about the brain. One of the most severe brain conditions, epilepsy is brought on by a group of
neurons in the brain that exhibits abnormally pathological oscillatory activity. Automated methods that
evaluate and identify epileptic episodes using electroencephalography data are currently being created.
This study's goal is to evaluate how well the ensemble approach-based model can foretell whether or not
an epileptic seizure would occur. The proposed mode was evaluated using the benchmark clinical dataset
provided by Bonn University. In this research, a reliable method based on temporal frequency image
(TFI) and fractal dimension(FD) is proposed for epilepsy detection -based EEG signals. The ensemble
method which consists of classifier KNN, SVM, and Tree is used to distinguish patients with epilepsy.
The outcomes demonstrated that the suggested strategy produces a high-accuracy clustering and aids
neurologists in making an epileptic condition diagnosis and subsequently recommending the proper
treatment. The suggested strategy is a promising one. approach for analyzing EEG signals and providing
reliable and accurate clustering to patients who suffer from epilepsy.

Published

2023-04-05