EEG-based single-channel authentication systems with optimum electrode placement for different mental activities.

Area of Science:

• Neuroscience
• Biometrics
• Signal Processing

Background:

• Electroencephalogram (EEG) signals possess unique patterns suitable for biometric authentication.
• Brainwave-based authentication offers advantages over existing biometric systems.
• Security and authentication are critical in daily life.

Purpose of the Study:

• To evaluate single-channel EEG-based authentication systems.
• To identify optimal EEG channels for authentication based on mental activities.

Main Methods:

• Utilized a dataset of five mental activities from seven subjects (325 samples).
• Employed pre-processing, feature extraction (Fourier, wavelet, autoregressive, entropy), and classification (Neural Network, Bayesian Network, SVM).
• Analyzed single-channel performance and channel optimization.

Main Results:

• Achieved 97-98% mean accuracy with a Neural Network classifier for single-channel authentication.
• Identified channel O2 as optimal for general authentication with 95% accuracy.
• Demonstrated that channel optimization improves performance and reduces channel count.

Conclusions:

• Channel optimization enhances EEG-based authentication performance.
• Optimal electrode placement can be defined for different mental activities.
• Single-channel systems with optimized channels show significant potential for authentication.