Dynamic complexity of cognitive event-related potential based on time-dependent approximate entropy

A method for calculating dynamic complexity of cognitive event-related potential (ERP) was introduced, which was recognized as time-dependent approximate entropy (ApEn). In addition, two calculating methods, mean of entropy for each trial (Mean of ApEn) and entropy for averaged ERP (ApEn of ERP), were validated and compared through an investigation of EEG complexity in a Chinese Stroop task containing two types of stimulus. Two phenomena can be observed from the experiment: 1. In contrast to the ApEn before stimulus, ApEn from two types of tasks decreased remarkably after the presence of stimulus, reached a minimum corresponding to the response output and rose back after response over frontal, central, and parietal regions. 2. ApEn of ERP had significant difference among three periods: pre-stimulus, within-response, and post-response, while Mean of ApEn did not. It is demonstrated that time-dependent approximate entropy is an effective method to measure the dynamic complexity of cognitive ERPs.