Abstract
Spikes are one of the main characteristics of a seizure electroencephalogram (EEG) signal. This feature plays an important role in seizure abnormality detection in EEG signals. The objective of this work is to provide a methodology to enhance this characteristic in the time-frequency domain. To achieve this goal first, we amplify the spike components in the raw EEG signal using the differential window, then a modified version of adaptive directional time-frequency distribution of the amplified signal is computed. The performance of the proposed method assessed using a simulated and a real EEG data. The results show an improvement in the time-frequency representations of a signal with spikes components. Different TFDs are tested, the modified-ADTFD provides the best performance among the selected TFDs.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call