Technique of automated hypnogram construction [Electronic resource] / E. S. Zakharov, P. P. Kravchenko, A. A. Skomorokhov
Уровень набора: (RuTPU)RU\TPU\book\169973, Bulletin of the Tomsk Polytechnic University / Tomsk Polytechnic University (TPU) = 2006-2007Язык: английский ; оригинала, русский.Страна: Россия.Описание: 1 файл (437 Кб)Серия: Control, computer engineeringand information scienceРезюме или реферат: The technique of automated sleep stage recognition and hypnogram construction has been considered. For partition of initial polysomnogram by segments obtained as a result of patient sleep monitoring the signal energy is analyzed using nonlinear energy controller. Frequency weighted energy is calculated for all registered signals then averaging and segmentation occur according to monitored signals behavior. Secondary index vector which is used at transition from segments to fixed duration periods is formed for segments. One or another sleep stage is finally assigned to the period by correlation analysis. Accuracy of the developed algorithm is connected with quantity of considered secondary indices, maximally detailed description of sleep stage characteristics and realization of training by manually prepared examples.Примечания о наличии в документе библиографии/указателя: [Bibliography: p. 126 (7 titles)].Тематика: hypnograms | automated recognition | sleep | segments | polysomnograms | monitoring | patients | signals | energy controllers | frequency weighted energy | vectors | correlation analysis | secondary indices | examples | электронный ресурс Ресурсы он-лайн:Щелкните здесь для доступа в онлайнЗаглавие с титульного листа
Электронная версия печатной публикации
[Bibliography: p. 126 (7 titles)]
The technique of automated sleep stage recognition and hypnogram construction has been considered. For partition of initial polysomnogram by segments obtained as a result of patient sleep monitoring the signal energy is analyzed using nonlinear energy controller. Frequency weighted energy is calculated for all registered signals then averaging and segmentation occur according to monitored signals behavior. Secondary index vector which is used at transition from segments to fixed duration periods is formed for segments. One or another sleep stage is finally assigned to the period by correlation analysis. Accuracy of the developed algorithm is connected with quantity of considered secondary indices, maximally detailed description of sleep stage characteristics and realization of training by manually prepared examples
Adobe Reader
Для данного заглавия нет комментариев.