Computer Scoririg of Motility Patterns for States of Sleep and Wakefulness: Human Infants
Author(s) -
Evelyn B. Thoman,
Rosemary C. Glazier
Publication year - 1987
Publication title -
sleep
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.222
H-Index - 207
eISSN - 1550-9109
pISSN - 0161-8105
DOI - 10.1093/sleep/10.2.122
Subject(s) - wakefulness , sleep (system call) , audiology , template matching , electroencephalography , rhythm , psychology , sleep stages , medicine , neuroscience , computer science , polysomnography , artificial intelligence , image (mathematics) , operating system
A procedure is described for continuous monitoring of sleep states and wakefulness of infants in the home, with no intervention required for the recording. The infant's motility patterns produced by respiration and body movements are recorded from a pressure-sensitive mattress in the crib onto a single channel of a small 24-h analog recorder. In this way, signals produced by the naturally occurring states are obtained during all portions of the day when the infant is in the crib. Later, in the laboratory, the signals are analyzed by computer, using a pattern recognition program with a template-matching procedure that scores the data every 30 s for Active Sleep, Quiet Sleep, Active-Quiet Transition Sleep, Sleep-Wake Transition, and Wakefulness. Computer-scored records for 2-h recordings of 10 infants were compared with direct behavioral state observations that were made at the same time as the mattress recording. The results confirmed the validity of the computer scoring procedure, both when templates were obtained from each infant's observation and when a standard set of templates from one infant were used for scoring the data from all other subjects.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom