| Abstract | A classification strategy of neonatal sleep is being developed by comparing
visually scored minutes of 21 channels of electroencephalographic
(EEG)/polygraphic recordings with the corresponding values for each
physiological signal derived from either visual or computer
analyses. Continuous 3-hour sleep studies on 54 preterm and full-term
neonates at similar postconceptional term ages were acquired under
environmentally controlled conditions using a computerized
monitoring system. An on-line event marker program recorded behavioral
observations. One of three EEG sleep states was assigned to each of 8,995
minutes by traditional visual analysis criteria. EEG spectral values,
spectral and nonspectral cardiorespiratory calculations and behaviorally
observed movements, arousals and rapid eye movement counts were submitted for
discriminant analysis. Based on the total minutes known for each of three
states (i.e. active, quiet and awake), linear combinations of all specified
digitized parameters were formed into an arithmetic algorithm by use of
discriminant analysis, which served as the basis of a state assignment for
each minute. Fifty percent of the data were arbitrarily used as the training
set to derive the state classification model. The remaining fifty percent of
the data were used as the cross-validation "test sample" to determine the
accuracy of the classification when compared to the visually analyzed score
for each corresponding minute. Thirteen out of 32 physiological measures best
predicted state of both preterm and full-term neonatal groups. For both
groups, the correct classification for active sleep was 90.3%, quiet sleep
was 97.4%, awake was 97% and the overall accuracy was 93.3%. However, the
order of significance for specific variables differed between these two
neonatal groups. Differences in the order of variables that predict sleep
states between preterm and full-term infants may reflect adaptation of brain
function of the preterm infant to prematurity and/or
prolonged extrauterine experience. |