Extracting epileptic features in EEGs using a dual-tree complex wavelet transform coupled with a classification algorithm

• Published in: Brain research 2022-03, Vol.1779, p.147777-147777, Article 147777
• Main Authors: Al-Salman, Wessam, Li, Yan, Wen, Peng, Miften, Firas Sabar, Oudah, Atheer Y., Ghayab, Hadi Ratham Al
• Format: Article
• Language: English
• Subjects: Adult; Algorithms; Cerebral Cortex - physiopathology; Dual-tree complex wavelet transform; Electroencephalography - methods; Electroencephalography - standards; Epilepsy - diagnosis; Epileptic seizures; Fast Fourier transformation; Focal, non-Focal EEG signals; Fourier Analysis; Humans; Seizures - diagnosis; Signal Processing, Computer-Assisted; Wavelet Analysis
• ISSN: 0006-8993; 1872-6240
• DOI: 10.1016/j.brainres.2022.147777

Electroencephalography (EEG), the most widely used technique for diagnosis, records the neuronal activity in the brain and thus the evaluation of epileptic seizures. Epilepsy is the second most common neurological disorder of the brain caused by abnormal excessive neuronal activity in the brain. The detection of epileptic seizures from electroencephalogram (EEG) signals is traditionally performed by clinical experts through visual inspection. It takes considerable time long duration to carry out an analysis so is time-consuming, requires more experience, and can be affected by the clinicians, subjectively. In this paper, EEG-based epileptic seizures detection is assessed by employing Bern–Barcelona EEG and the Bonn University EEG database. The proposed method contains three steps: segmentation techniques, features extraction, and classification. Initially, each EEG signal is divided into four segments and sub-segments, and then each segment is further split into smaller sub-segments. After that, hybrid transformation techniques based on DTCWT and FFT are improved in order to extract robust features from each sub-segment. Finally, a set of effective features are extracted from the sub-segments and are then forwarded to the LS-SVM classifier to classify epileptic EEGs. The experimental results demonstrate that the method obtains an average accuracy of 97.7% and 96.8% for the Bonn and Bern databases, respectively. This study suggests that the proposed method could be an effective features extraction technique for the classification of epileptic EEG signals, and it can be adapted to aid neurologists to better diagnose neurological disorders and therefore for an early seizure warning system. [Display omitted] •DT-CWT and classification algorithm are used to identify epileptic seizures.•The DT-CWT and FFT both are utilized to extract the EEG features.•The LS-SVM classifier is used to detect the epileptic patterns in EEG.•This method provides better detection accuracy compared to other existing approaches.•The obtained results showed that the proposed method outperformed the others. The detection of epileptic seizures from electroencephalogram (EEG) signals is traditionally performed by clinical experts through visual inspection. It is a long process, is error prone, and requires a highly trained expert. In this research, a new method is presented for seizure classification for EEG signals using a dual-tree complex wavelet transform (DT-CWT) and fast Fourier tran