TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

Unobtrusive Monitoring of Heart Rate and Respiration Rate during Sleep

Tutkimustuotos

Yksityiskohdat

AlkuperäiskieliEnglanti
KustantajaTampere University of Technology
Sivumäärä70
ISBN (elektroninen)978-952-15-3475-1
ISBN (painettu)978-952-15-3470-6
TilaJulkaistu - 27 helmikuuta 2015
OKM-julkaisutyyppiG5 Artikkeliväitöskirja

Julkaisusarja

NimiTampere University of Technology. Publication
KustantajaTampere University of Technology
Vuosikerta1285
ISSN (painettu)1459-2045

Tiivistelmä

Sleep deprivation has various adverse psychological and physiological effects. The effects range from decreased vigilance causing an increased risk of e.g. traffic accidents to a decreased immune response causing an increased risk of falling ill. Prevalence of the most common sleep disorder, insomnia can be, depending on the study, as high as 30 % in adult population. Physiological information measured unobtrusively during sleep can be used to assess the quantity and the quality of sleep by detecting sleeping patterns and possible sleep disorders. The parameters derived from the signals measured with unobtrusive sensors may include all or some of the following: heartbeat intervals, respiration cycle lengths, and movements. The information can be used in wellness applications that include self-monitoring of the sleep quality or it can also be used for the screening of sleep disorders and in following-up of the effect of a medical treatment. Unobtrusive sensors do not cause excessive discomfort or inconvenience to the user and are thus suitable for long-term monitoring. Even though the monitoring itself does not solve the sleeping problems, it can encourage the users to pay more attention on their sleep. While unobtrusive sensors are convenient to use, their common drawback is that the quality of the signals they produce is not as good as with conventional measurement methods. Movement artifacts, for example, can make the detection of the heartbeat intervals and respiration impossible. The accuracy and the availability of the physiological information extracted from the signals however depend on the measurement principle and the signal analysis methods used. Three different measurement systems were constructed in the studies included in the thesis and signal processing methods were developed for detecting heartbeat intervals and respiration cycle lengths from the measured signals. The performance of the measurement systems and the signal analysis methods were evaluated separately for each system with healthy young adult subjects. The detection of physiological information with the three systems was based on the measurement of ballistocardiographic and respiration movement signals with force sensors placed under the bedposts, the measurement of electrocardiographic (ECG) signal with textile electrodes attached to the bed sheet, and the measurement of the ECG signal with non-contact capacitive electrodes. Combining the information produced by different measurement methods for improving the detection performance was also tested. From the evaluated methods, the most accurate heartbeat interval information was obtained with contact electrodes attached to the bed sheet. The same method also provided the highest heart rate detection coverage. This monitoring method, however, has a limitation that it requires a naked upper body, which is not necessarily acceptable for everyone. For respiration cycle length detection, better results were achieved by using signals recorded with force sensors placed under a bedpost than when extracting the respiration information from the ECG signal recorded with textile bed sheet electrodes. From the data quality point of view, an ideal night-time physiological monitoring system would include a contact ECG measurement for the heart rate monitoring and force sensors for the respiration monitoring. The force sensor signals could also be used for movement detection.

Tilastokeskuksen tieteenalat

Latausten tilastot

Ei tietoja saatavilla