Related Experiment Videos
Autoregressive modeling of epicardial electrograms during ventricular fibrillation
R Throne1, D Wilber, B Olshansky
1Department of Electrical Engineering, University of Nebraska, Lincoln 68588.
IEEE Transactions on Bio-Medical Engineering
|April 1, 1993
Summary
This study modeled ventricular fibrillation (VF) electrograms using autoregressive processes. Synthesized VF signals closely matched real ones, suggesting potential for improved defibrillator algorithm testing.
Area of Science:
- Biomedical Engineering
- Signal Processing
- Cardiology
Background:
- Ventricular fibrillation (VF) electrograms typically lack discernible organization.
- Automatic implantable defibrillators rely on analyzing these signals for effective treatment.
Purpose of the Study:
- To determine if VF electrograms can be modeled as an autoregressive stochastic process.
- To synthesize VF signals for enhanced testing of defibrillator algorithms.
Main Methods:
- Modeled VF electrograms from bipolar epicardial electrodes using an autoregressive process with white noise excitation.
- Synthesized ten VF signals for each patient using the derived model.
- Compared synthesized and original VF signals using metrics like rms amplitude, zero crossings, and rate variation.
Main Results:
- Synthesized VF waveforms exhibited similar root mean square (rms) amplitudes to true VF.
- Differences in rate, RR interval regularity, zero crossings, and baseline time between synthesized and true VF signals were generally not statistically significant (p >= 0.05).
Conclusions:
- VF electrograms can be modeled as autoregressive stochastic processes.
- Synthesized VF signals demonstrate sufficient similarity to real signals.
- This approach may enable more comprehensive testing of VF detection algorithms.