Toward computerized scoring of sleep using ambulatory recordings of heart rate and physical activity

Kent F. Burnett, C. Barr Taylor, Carl E. Thoresen, Mark R. Rosekind, Laughton E. Miles, Robert F. DeBusk

Research output: Contribution to journalArticlepeer-review

3 Scopus citations


A microprocessor-based ambulatory monitoring system for recording heart rate and physical activity level was used to assess sleep/wake patterns in the natural environment. Two normal subjects and eight subjects who complained of disturbed sleep wore the monitoring system continuously over a 60-hour period while undergoing two all-night home polysomnographic recordings of EEG, EOG, and EMG. This study reports the degree of correspondence obtained when sleep parameters determined from the polysomnographic recordings were compared to those determined by subjective scoring of computer generated graphs which depicted heart rate and physical activity level over time, and by a computer program which used a statistical model to classify the data into sleep versus wake categories automatically. The data showed a high degree of consistency between all methods for measuring latency to final wake. A high degree of consistency also was found between sleep onset times determined by polysomnography and sleep onset times determined from subjective scoring of the heart rate and physical activity graphs. Possible methods for deriving computer-assisted estimates for other sleep parameters such as REM, wake after sleep onset and total sleep time are discussed. Microprocessor-based ambulatory recording combined with computer-assisted scoring shows great potential for use in sleep research.

Original languageEnglish (US)
Pages (from-to)261-271
Number of pages11
JournalBehavioral Assessment
Issue number3
StatePublished - Jun 1 1985

ASJC Scopus subject areas

  • Psychiatry and Mental health


Dive into the research topics of 'Toward computerized scoring of sleep using ambulatory recordings of heart rate and physical activity'. Together they form a unique fingerprint.

Cite this