Scalp electroencephalography（ EEG） signals reflects the spontaneous rhythmic electrical physiological activities of cerebral cortex neurons

which are rich in physiological and pathological information

and important for clinical neurological and mental disease diagnosis. Aiming at condition that the depression study and diagnosis are lack of deffective quantitative parameters and quantitative indicators

present a new EEG signal analysis method based on wavelet packet decomposition

power spectrum entropy

calculate the power spectral entropy value

W value of node reconstruction signal power spectrum. Apply this metho

d to calculate and analyze the resting EEG signals. The results show that the W value of S32 wavelet packet node with frequency range of 24

32 Hz with depression patients are significantly larger than normal healthy people in some brain areas. It indicates that the distribution of fast EEG rhythm wave energy is more complicated and dispersive in depression patients

confirming to the pathogenesis about weak self-regulated behavior about depression patients. T-test statistical analysis is performed to prove the feasibility and accuracy of this method

provide an objective and effective quantitative physical feature indicator for depression disease diagnosis.