We assess the complexity of Heart Rate Variability (HRV) time series by computing Approximate Entropy (ApEn) and Sample Entropy (SampEn) on different time scales, according to a the Multiscale Entropy method (MSE) recently introduced. We find that considering multiple time scales enhances the ability to differentiate time series obtained from healthy and pathological subjects. Simulated signals such as white noise and 1/f^alpha noise have been analyzed to suggest possible reasons for these differences in HRV. Data belong to 10 Normal subjects, and patients who suffered from Myocardial Infarction (10), Congestive Heart Failure (10) and Heart transplant (10), all collected over 24 hours. Results show that the calculated parameters significantly separate the patient classes. The best performance is obtained for scale factors 3 to 8. We believe that the results could be exploited in order to reinforce indicators of risk prediction in cardiovascular pathologies.
Multiscale entropy analysis of 24 hours heart rate variability time series / M. Ferrario, M.G. Signorini, R. Sassi, S. Cerutti (IFMBE PROCEEDINGS). - In: Medicon and health telematics : proceeding / [a cura di] M. Bracale. - [s.l] : IFMBE, 2004. - ISBN 8877803088. (( Intervento presentato al 10. convegno Mediterranean Conference on Medical and Biological Engineering and Computing tenutosi a Ischia nel 2004.
Multiscale entropy analysis of 24 hours heart rate variability time series
R. SassiPenultimo
;
2004
Abstract
We assess the complexity of Heart Rate Variability (HRV) time series by computing Approximate Entropy (ApEn) and Sample Entropy (SampEn) on different time scales, according to a the Multiscale Entropy method (MSE) recently introduced. We find that considering multiple time scales enhances the ability to differentiate time series obtained from healthy and pathological subjects. Simulated signals such as white noise and 1/f^alpha noise have been analyzed to suggest possible reasons for these differences in HRV. Data belong to 10 Normal subjects, and patients who suffered from Myocardial Infarction (10), Congestive Heart Failure (10) and Heart transplant (10), all collected over 24 hours. Results show that the calculated parameters significantly separate the patient classes. The best performance is obtained for scale factors 3 to 8. We believe that the results could be exploited in order to reinforce indicators of risk prediction in cardiovascular pathologies.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
