Developmental changes in electroencephalogram for term and preterm infants using an autoregressive model

The purpose of the present study was to clarify the developmental changes of electroencephalogram (EEG) background activity in term and preterm infants that take place with increasing conceptional age (CA) through an autoregressive (AR) model. Polygraphical EEG recordings were obtained from 76 healthy term and preterm infants with a CA ranging between 31 and 40 weeks. Decreases in total power and component power of δ were noted with CA during burst activity (BA) in quiet sleep (QS) and during active sleep (AS). Increases in total power and component power of δ were noted with CA during interburst interval activity (IBIA) of QS. Regression analysis (RA) of the information amount (IA) indicated a negative correlation with increasing CA in AS and in the monopolar EEG tracings Fp1, C3, O1 and O2 in BA, and a positive correlation in O1 in IBIA. Regression analysis of the IA of δ indicated a negative correlation with increasing CA in AS and in Fp1, C3, O1 and O2 in BA, and a positive correlation in O1 in IBIA. Regression analysis of the IA of δ indicated a negative correlation with increasing CA in BA and AS. These results showed that the high voltage slow wave component changed to a low voltage slow one with development in AS and that BA was longer in duration and lower in power with increasing CA, while IBIA was shorter in duration and higher in power. In conclusion, significant developmental changes occur in all derivatives of AS. Even though the EEG of BA and IBIA change separately, they are followed by EEG of QS in a continuous pattern.