Single trial analysis on EEG signatures to identify individuals

This paper presents a biometric authentication system using brain signature to identify individuals. Research on brain signals has shown that each individual has a unique brain wave pattern for similar activities. EEG signals are recorded for three mental tasks from different subjects to extract the distinctive brain signature of the individual. 15 subjects participated in this experiment. A novel protocol and algorithm using neural networks is proposed. Single trial analysis is performed to test the performance of the proposed algorithm. The identification performance of the neural network by single trial analysis is 78.6%.