سامانه همایش ها

صفحه اصلی / سی و دومین کنفرانس ملی و دهمین کنفرانس بین المللی مهندسی زیست پزشکی ایران

Comparative Analysis of Machine Learning and Deep Learning Models for Epileptic Seizure Detection Using the CHB-MIT EEG Dataset

نویسندگان :

Pouya Taghipour Langrodi¹ Amirsadra Khodadadi² Mahtab Dastranj³ Golnaz Baghdadi⁴

1- Department of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran 2- Department of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran 3- Department of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran 4- Department of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran

کلمات کلیدی :

Epilepsy،Neural Networks،Seizure Detection،Electroencephalography،EEG،Deep Learning،Machine Learning،LSTM

چکیده :

Epilepsy is one of the most common neurological disorders that usually comes with sudden and unpredictable seizures and can severely affect the quality of life of patients. This study aims to design and evaluate different artificial approaches for automated seizure detection using EEG signals from the CHB-MIT dataset. This dataset contains 23 patients suffering from epileptic seizures, including boys and girls aged between 1.2 to 22 years old. Feature extraction was performed across time, frequency, and time-frequency domains. Eight classifiers were implemented in this study, including four machine learning algorithms (SVM, KNN, Decision Tree, and naïve Bayes) and four deep learning architectures (Artificial Neural Network, LSTM, TCN, and Transformer). The results demonstrated that the LSTM and TCN models outperformed other classifiers in detecting the preictal and ictal stages, achieving an accuracy of 96.0% and 97.3% with the sensitivity of 93.5% and 90.5%. Moreover, ANN and Transformer achieved 94.8% and 93.2% accuracy. In contrast, SVM, KNN, DT, and NB represented 93.1%, 92.4%, 81.2%, and 71.9% in accuracy. By preparing a uniform data preparation baseline for the CHB-MIT dataset, this study made an identical comparison between machine learning and deep learning models to propose the best approach for epileptic seizure detection.