Abstract
Most physiological time series have self-similar properties which reflect the functioning of physiological control mechanisms. Self-similarity is usually assessed by detrended fluctuation analysis (DFA) assuming that mono- or bi-fractal models generate the self-similar dynamics. Our group recently proposed a new DFA approach describing self-similarity as a continuous temporal spectrum of coefficients, thus not assuming that "lumped-parameter" fractal models generate the data. This paper reviews the rationale for calculating a spectrum of DFA coefficients and applies this method on datasets of signals whose self-similarity has been extensively studied in the past. The first dataset consists of six electroencephalographic (EEG) derivations collected in a healthy volunteer. The second dataset consists of cardiac intervals and diastolic blood pressures recorded in 60 volunteers at different levels of cardiac sympatho/vagal balance. Results reveal the limits of the traditional "lumped-parameter" approach, and provide information on the role of autonomic outflows in determining cardiovascular self-similarity.
| Original language | English |
|---|---|
| Title of host publication | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
| Pages | 164-175 |
| Number of pages | 12 |
| Volume | 7548 LNBI |
| DOIs | |
| Publication status | Published - 2012 |
| Event | 8th International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics, CIBB 2011 - Gargnano del Garda, Italy Duration: Jun 30 2011 → Jul 2 2011 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 7548 LNBI |
| ISSN (Print) | 03029743 |
| ISSN (Electronic) | 16113349 |
Other
| Other | 8th International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics, CIBB 2011 |
|---|---|
| Country | Italy |
| City | Gargnano del Garda |
| Period | 6/30/11 → 7/2/11 |
Fingerprint
Keywords
- Detrended fluctuation analysis
- EEG
- Fractals
- Heart rate variability
ASJC Scopus subject areas
- Computer Science(all)
- Theoretical Computer Science
Cite this
Self-similarity in physiological time series : New perspectives from the temporal spectrum of scale exponents. / Castiglioni, Paolo.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7548 LNBI 2012. p. 164-175 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7548 LNBI).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Self-similarity in physiological time series
T2 - New perspectives from the temporal spectrum of scale exponents
AU - Castiglioni, Paolo
PY - 2012
Y1 - 2012
N2 - Most physiological time series have self-similar properties which reflect the functioning of physiological control mechanisms. Self-similarity is usually assessed by detrended fluctuation analysis (DFA) assuming that mono- or bi-fractal models generate the self-similar dynamics. Our group recently proposed a new DFA approach describing self-similarity as a continuous temporal spectrum of coefficients, thus not assuming that "lumped-parameter" fractal models generate the data. This paper reviews the rationale for calculating a spectrum of DFA coefficients and applies this method on datasets of signals whose self-similarity has been extensively studied in the past. The first dataset consists of six electroencephalographic (EEG) derivations collected in a healthy volunteer. The second dataset consists of cardiac intervals and diastolic blood pressures recorded in 60 volunteers at different levels of cardiac sympatho/vagal balance. Results reveal the limits of the traditional "lumped-parameter" approach, and provide information on the role of autonomic outflows in determining cardiovascular self-similarity.
AB - Most physiological time series have self-similar properties which reflect the functioning of physiological control mechanisms. Self-similarity is usually assessed by detrended fluctuation analysis (DFA) assuming that mono- or bi-fractal models generate the self-similar dynamics. Our group recently proposed a new DFA approach describing self-similarity as a continuous temporal spectrum of coefficients, thus not assuming that "lumped-parameter" fractal models generate the data. This paper reviews the rationale for calculating a spectrum of DFA coefficients and applies this method on datasets of signals whose self-similarity has been extensively studied in the past. The first dataset consists of six electroencephalographic (EEG) derivations collected in a healthy volunteer. The second dataset consists of cardiac intervals and diastolic blood pressures recorded in 60 volunteers at different levels of cardiac sympatho/vagal balance. Results reveal the limits of the traditional "lumped-parameter" approach, and provide information on the role of autonomic outflows in determining cardiovascular self-similarity.
KW - Detrended fluctuation analysis
KW - EEG
KW - Fractals
KW - Heart rate variability
UR - http://www.scopus.com/inward/record.url?scp=84871553910&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84871553910&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-35686-5_14
DO - 10.1007/978-3-642-35686-5_14
M3 - Conference contribution
SN - 9783642356858
VL - 7548 LNBI
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 164
EP - 175
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ER -