Multichannel EEG based inter-ictal seizures detection using Teager energy with backpropagation neural network...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Epileptic seizure detection from long-term electroencephalogram (EEG) recordings is critical for patient care.
• Automated detection methods are needed to efficiently analyze complex EEG data.

Purpose of the Study:

• To investigate the automated detection of epileptic seizures from multichannel EEG data.
• To evaluate the efficacy of the Teager energy feature and a back-propagation neural network for seizure classification.

Main Methods:

• Utilized multichannel EEG data from a hospital setting after ethical approval.
• Applied signal processing techniques including notch filtering and independent component analysis.
• Extracted Teager energy features from 1-second windowed EEG segments.
• Implemented a supervised back-propagation neural network for inter-ictal seizure classification.

Main Results:

• Teager energy feature demonstrated suitability for seizure detection through descriptive and box plot analysis.
• The neural network classifier achieved high performance metrics.
• Achieved a sensitivity of 96.66%, specificity of 99.15%, and a false detection rate of 0.30 per hour.

Conclusions:

• The proposed automated seizure detection procedure using Teager energy and neural networks is effective.
• The method shows promise for real-time epileptic seizure detection applications.