logo
 

logo
logo


Достовірність методу визначення розладки ритмокардіосигналу

НазваДостовірність методу визначення розладки ритмокардіосигналу
Назва англійськоюReliability of method for change-point detection of rhythm–cardiosignal
АвториЯворський Б. Лещишин Ю.
Бібліографічний описЯворський Б. Достовірність методу визначення розладки ритмокардіосигналу / Б. Яворський, Ю. Лещишин // Вісник ТНТУ — Тернопіль : ТНТУ, 2014. — Том 73. — № 1. — С 252-258. — (математичне моделювання. математика. фізика).
Bibliographic description:Yavorskyy B. Reliability of method for change-point detection of rhythm–cardiosignal / B. Yavorskyy, Yu. Leschyshyn // Bulletin of TNTU — Ternopil : TNTU, 2014. — Volume 73. — No 1. — P 252-258. — (mathematical modeling. mathematics. physics).
УДК:

621.391.7
612.172.2

Ключові слова

ритмокардіосигнал
визначення розладки
достовірність
періодично корельована випадкова послідовність
критерій Неймана- Пірсона
rhythmocardiosignal
change-point detection
reliability
periodically correlated stochastic sequence
the Neyman-Pearson criterion

Отримано характеристики достовірності методів визначення розладки ритмокардіосигналу. Обґрунтовано вибір критерію ефективності відповідного методу. Проведено порівняння достовірності методів визначення розладки ритмокардіосигналу, тестова статистика яких отримана синфазним методом, методом періодограм, методом Берга, методом Уелча.
While analyzing the heart rate variability the sequence of time intervals between RR teeth – rhythmocardiosignal is analyzed with the help of the methods based on its stationary model. Different changes in the physical activity, physiological state of a human and environmental factors cause sufficient changes in the statistical characteristics of rhythmocardiosignal – their change-point and nonstationarity, which results in the decrease of the reliability of the rhythmocardiosignal variability parameters estimation. Thus, the mathematical model of rhythmocardiosignal, which varies from stationary to periodically correlated stochastic sequence, was built to take into account the rhythmocardiosignal change-point appearance. The method of change-point detection of rhythmocardiosignal will be based on the comparison with a threshold of the likelihood ratio test statistic. The threshold of change-point detection of rhythmocardiosignal is obtained due to the Neumann-Pearson criterion, when the probability of false change-point detection is set. The test statistic of the change-point detection of rhythmocardiosignal is obtained taking advantage of the methods of spectral analysis, the synphase method, the periodogram method, the Berg and Welch methods in particular. The reliability of decisions on the appearance of the change-point of rhythmocardiosignal, ie the probability of the correct detection of the change-point, which accounts the statistical nature of the appearance of the rhythmocardiosignal change-point, was used as the criterion of methods efficiency of the change-point detection of rhythmocardiosignal. The characteristics of the reliability of the methods of change-point detection of rhythmocardiosignal, which were selected as the criterion of their efficiency, have been obtained. The comparison of the reliability of the methods of the change-point detection of rhythmocardiosignal, the test statistics of which was obtained taking advantage of the synphase method, the periodogram method, the Berg method, the Welch method, has been carried out. The obtained results testify that the reliability of change-point detection of rhythmocardiosignal by the synphase method is in 2 times higher then that obtained by the periodogram method, in to 1,8 times by the Berg method and in 1,4 times by the Welch method, the probability of a false change-point detection being up to 0,1 and relation to 4dB of the nonstationary energy to power spectral density of the stationary rhythmocardiosignal.

ISSN:1727-7108
Завантажити

Всі права захищено © 2016. Тернопільський національний технічний університет імені Івана Пулюя.