Title :
Morphological descriptors based on Eigen value decomposition for P-wave analysis
Author :
Castells, F. ; Lorenz, J. ; Climent, AM ; Guillem, Maria S. ; Husser, D. ; Bollmann, A. ; Millet, J.
Author_Institution :
Univ. Politec. de Valencia, Valencia
Abstract :
In this study, several morphological parameters to assess the P-wave from the surface electrocardiogram are proposed and analysed. In order to evaluate spatial features, two parameters are defined: P-wave complexity and P-wave residuum. Additionally, in order to evaluate beat to beat features, the P-wave regularity is defined. All three parameters are computed from the eigenvalue decomposition of the original data, and were applied on patients who suffered from either dilated or ischemic cardiomyopathy. The results showed that both the P-wave complexity and P-wave regularity are able to find and emphasize interpatient differences, whereas the results of P-wave residuum where rather similar for all patients.
Keywords :
diseases; eigenvalues and eigenfunctions; electrocardiography; medical signal processing; ECG signal preprocessing; P-wave analysis; P-wave complexity; P-wave regularity; P-wave residuum; eigen value decomposition; ischemic cardiomyopathy; morphological descriptor; spatial features; surface electrocardiogram; Cardiology; Data mining; Eigenvalues and eigenfunctions; Electrocardiography; Feature extraction; Rhythm; Risk analysis; Robustness; Surface morphology; Surgery;
Conference_Titel :
Computers in Cardiology, 2008
Conference_Location :
Bologna
Print_ISBN :
978-1-4244-3706-1
DOI :
10.1109/CIC.2008.4749023