Comparison Between Time‐Lapse Video Recordings of Behavior and Polygraphic State Determinations in Premature Infants

Abstract

This study was done to compare time‐lapse video recordings of eye and body activity with polygraphically determined states of activity from human premature infants. Time‐lapse video recording in neonates is new, and was first reported by Anders and Sostek (1976) who have used time‐lapse video recordings of sleep‐wake activity in full‐term infants. We recorded 8 prematurely born human infants, 5 males and 3 females between 29 and 34 weeks gestational age at birth. Behavioral and physiological states were scored from the polygraphic records. The video‐tape observations were coded for three levels of body movement: no body movement, some body movement, and a lot of body movement, and eyes: closed no rapid eye movements (REMs), REMs, and open. The following relationships (concordance) were observed: 1) 60% of the non‐REM/no movement epochs were in the Quiet state on the polygraph; 2) 83% of the REM/some movement epochs were in the REM state on the polygraph; and 3) 82% of the REM/ a lot of movement epochs were in the REM state on the polygraph.

Healthy Preterm Infant Responses to Taped Maternal Voice

“…44 This categorical scale was developed from multiple hours of videotaping of preterm infants and ranges from no facial or body activity to intense trunk movement (writhing or thrashing). The scale was validated by comparing time-lapsed video recordings of behavior and polygraphic state determinations on infants from 29 to 34 weeks' gestation, 45 and it has been used in other studies examining activity levels of premature infants. 44 These activity categories, along with the behaviors defined by Als, 41 were adapted to develop a computer generated coding scheme that assigned numerical values to each defined behavior, because it was selected to quantify the observed behaviors for analysis.…”

Sleep: A behavioral account.

“…Typically, the recording is viewed time compressed (i.e., viewed at a faster speed than recorded), so that time spent coding the record is reduced, but no automatic algorithm is yet available to obviate manual scoring by trained coders. Various scoring systems can be used to derive measures of sleep such as time spent in bed, time in active-REM and quiet-NREM sleep, time awake, and so forth (Fuller, Wenner, & Blackburn, 1978), and interobserver agreement is generally high, as is agreement with PSG (Anders & Sostek, 1976;Fuller et al, 1978). An audio channel is also available, and vocalizations, along with gestures, other activities, and parental or sibling interactions can all be detected and coded.…”

Audio- and video-based estimation of the sleep stages of newborns in Neonatal Intensive Care Unit

“…Fuller et al proposed a similar approach with premature newborns, but only focused on the eye state and the body movements. Furthermore, only sleep stages were considered and vocalizations were not included [10].…”