Title :
Body Surface ECG Signal Shape Dispersion
Author :
Khaddoumi, B. ; Rix, H. ; Meste, O. ; Fereniec, M. ; Maniewski, R.
Author_Institution :
Lab. of Informatics, Signals & Syst., Univ. of Nice, Sophia Antipolis
Abstract :
The spatial distribution of the shape of the electrocardiography (ECG) waves obtained by body surface potential mapping (BSPM) is studied, using a 64-channel high-resolution ECG system. The index associated to each lead is the shape difference between its ECG wave and a reference computed taking into account all the leads on the same column. The reference is either a selected real wave or a synthetic signal computed by integral shape averaging (ISA). Better results are obtained with the ISA signal using the distribution function method (DFM) for computing the shape difference. The spatial dispersion of ECG waves is showed to allow the separation of patients after myocardial infarction (MI) from healthy subjects. In addition, the reference signal position for each column is computed. The path linking these positions appears as an invariant, i.e., it is independent of the subject and the ECG wave
Keywords :
bioelectric potentials; electrocardiography; medical signal processing; body surface ECG signal shape dispersion; body surface potential mapping; distribution function method; electrocardiography; integral shape averaging; myocardial infarction; spatial distribution; Associate members; Design for manufacture; Distributed computing; Distribution functions; Electrocardiography; Heart rate interval; Instruction sets; Principal component analysis; Shape measurement; Surface waves; BSPM; ECG waves; shape analysis; spatial dispersion of shapes; Algorithms; Body Surface Potential Mapping; Computer Simulation; Diagnosis, Computer-Assisted; Electrocardiography; Humans; Models, Cardiovascular; Myocardial Infarction; Reproducibility of Results; Sensitivity and Specificity;
Journal_Title :
Biomedical Engineering, IEEE Transactions on
DOI :
10.1109/TBME.2006.881785
