چاپ صفحه |  صفحه نخست سامانه |  نویسندگان |  اطلاعات تفضیلی |  دانلود مقاله | دانشگاه علوم پزشکی تبریز |
| نویسنده | نفر چندم مقاله |
|---|---|
| سعید چارسوئی | چهارم |
| علی احمدعلی پور | پنجم |
| رضا جوانشیر | اول |
| محمدرضا صدقی گمچی | سوم |
| مهداد اسمعیلی | دوم |
| عنوان | متن |
|---|---|
| خلاصه مقاله | Epilepsy as a neurological disorder is usually diagnosed by manual interpretation of electroencephalography (EEG) by clinicians. An important problem in epilepsy is the classification of focal and non-focal EEG signals where the number of research works is limited and commonly it is needed for long-duration EEG signal acquisition containing seizures. The visual analysis of long-term EEG recordings is characterized by its subjectivity, time-consuming procedure, and erroneous detection. In this paper, we proposed an automated classification of resting-state EEG signals of epileptic patients into two focal and non-focal groups using learning algorithms. 50 subjects were selected for each group of patients with focal epilepsy (age range: 15-50, average: 27 years old) and non-focal epilepsy (age range: 15-50, average: 33 years old). Resting-state with eyes-closed epochs of both groups of EEG signals were extracted and analyzed. Feature extraction as the core trouble of brain signal processing performed by extracting kurtosis and energy features of the time-frequency domain of the EEG signals. These features help the classifiers to achieve good accuracy when used to automatically separate the EEG signal into two types of epilepsy. Several traditional machine learning and neural network classifiers have been hired to classify the EEG signal using the extracted features, which random forest performs best with an accuracy of 73.3%. Overall, the proposed classification method reports a significant improvement in the diagnoses of focal and non-focal epileptic EEG signals by using resting-state EEG signals in order to use long-duration signals containing seizures. Besides, it is very fast in comparison to the visual diagnose of the clinicians and can reduce the number of diagnoses errors. |
| کلمات کلیدی | EEG signal,Epilepsy,classification |
| نام فایل | تاریخ درج فایل | اندازه فایل | دانلود |
|---|---|---|---|
| Bernstein conference 2021.png | 1401/04/07 | 159352 | دانلود |
| poster-javanshir.pdf | 1401/04/07 | 234669 | دانلود |
| javanshir.png | 1401/04/07 | 22908 | دانلود |
