More normal EEGs of depression patients during mental arithmetic than rest

The aim of this study is to compare the brain activities in depression patients and healthy controls by a quantitative method. In the present study, the wavelet entropy (WE) and subband segmentation analysis are proposed to characterize the degree of disorder and complexity of the electroencephalographic (EEG) signals recorded from 16 scalp electrodes in 20 depression patients and 20 normal controls at rest and during mental arithmetic. The WE analysis demonstrates the EEGs of the depression subjects have higher WE than those of the controls at rest, which indicates a less rhythmic and ordered status in depression, whereas such difference is not significant during mental arithmetic. These results provide evidence that depression patients have more regular brain wave during mental arithmetic than rest. Furthermore, the WE of the depression patients is higher during rest than mental arithmetic at almost all electrodes, and this phenomenon is not found in the controls. It may suggest that larger parts of the brain of the patients are active during rest than performing a cognitive task. In addition, marginal effect of hemisphere is detected for the patients during mental arithmetic. Our results show that the we could be a useful tool in cognitive process analysis, especially in depression.