EEG analysis method for discrimination of positive / negative emotional state

摘要

The present invention relates to an EEG analysis method for determining a positive / negative emotional state, and to the EEG detected to determine the emotional positive / negative state of the subject through the analysis of the EEG without setting the baseline through one electrode attached to the scalp After the fast Fourier transform, the frequency band is divided into subbands, the relative power is calculated, and iteratively repeated to obtain the time zone relative power for the frequency components associated with the emotional processing occurring spontaneously in the subject's brain. In order to ignore fluctuations in time, it is relaxed by the average with neighboring data, and by using the least-squares method for the relaxed data, the gradient is increased over time, and the threshold is applied to the obtained gradient. Decrease the increase or decrease by determining the increase or decrease over time When the frequency component decreases and the high frequency component increases, it is determined as a negative emotional state, and when the low frequency component increases and the high frequency component decreases, it is determined as a positive emotional state, thereby enabling real-time use and It is possible to determine the positive / negative emotional state of the subject without setting the baseline by determining the positive / negative emotional state only by increase and decrease over time, and to minimize the inconvenience of the user when using it as a human-machine interface. Have