Measuring brain activity cycling (BAC) in long term EEG monitoring of preterm babies

Physiol Meas. 2014 Jul;35(7):1493-508. doi: 10.1088/0967-3334/35/7/1493. Epub 2014 Jun 5.


Measuring fluctuation of vigilance states in early preterm infants undergoing long term intensive care holds promise for monitoring their neurological well-being. There is currently, however, neither objective nor quantitative methods available for this purpose in a research or clinical environment. The aim of this proof-of-concept study was, therefore, to develop quantitative measures of the fluctuation in vigilance states or brain activity cycling (BAC) in early preterm infants. The proposed measures of BAC were summary statistics computed on a frequency domain representation of the proportional duration of spontaneous activity transients (SAT%) calculated from electroencephalograph (EEG) recordings. Eighteen combinations of three statistics and six frequency domain representations were compared to a visual interpretation of cycling in the SAT% signal. Three high performing measures (band energy/periodogram: R = 0.809, relative band energy/nonstationary frequency marginal: R = 0.711, g-statistic/nonstationary frequency marginal: R = 0.638) were then compared to a grading of sleep wake cycling based on the visual interpretation of the amplitude-integrated EEG trend. These measures of BAC are conceptually straightforward, correlate well with the visual scores of BAC and sleep wake cycling, are robust enough to cope with the technically compromised monitoring data available in intensive care units, and are recommended for further validation in prospective studies.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Artifacts
  • Brain / physiology*
  • Databases, Factual
  • Electroencephalography / methods*
  • Humans
  • Infant, Extremely Premature / physiology
  • Infant, Newborn
  • Infant, Premature / physiology*
  • Intensive Care, Neonatal
  • Linear Models
  • Nonlinear Dynamics
  • Signal Processing, Computer-Assisted
  • Time Factors