Regularity and Complexity of Human Electroencephalogram Dynamics: Applications to Diagnosis of Alzheimers Disease

In this paper, we evaluate the complexity and regularity of human electroencephalogram (EEG) dynamics using approximate entropy (ApEn), and the results are used to distinguish Alzheimer´s disease (AD) patients from healthy subjects. From the 10-channel EEG time series recordings of 20 healthy subjects and 14 AD patients with closed eyes, our analysis has shown that AD patients have lower ApEn values than healthy subjects. These results support the previous hypothesis that greater regularity corresponds to greater component autonomy and isolation in many complex systems. We believe that our effort provides a valuable complementary framework to the classical EEG analysis, and it could help revealing the complexity of the human brain functions