Year: 2026
Type: conference-paper
Abstract: The Global population prevalence of epilepsy is in millions requiring timely and accurate seizure classification for early management. The objective of this study is to apply machine learning (ML) algorithms to classify four types of EEG data: Complex Partial Seizures, Electro-graphic Seizures, Video detected Seizures and Normal data. The data collected for the study consists of EEG records of 6 epileptic patients with 21 channels sampled for 500 Hz. We considered five ML models, i.e. Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), CNN-LSTM, Time-Distributed CNN, and Support Vector Machine (SVM). The models were trained from 90% of the data and tested on the remaining 10%. The study shows that CNN-LSTM model has the highest level of accuracy 89% for the classification of seizures in comparison to other traditional and deep learning models. Evaluation parameters including accuracy, precision, recall and F1 score which were used on all the models. These findings indicate that a combination of CNN and LSTM can capture temporal and spatial patterns of EEG data for improving Seizure classification accuracy. In doing so, this research enhances the existing body of work on automated analysis of EEG data for epilepsy diagnosis in clinical practice.
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: diamond
OpenAlex ID: W7153296342