Related Experiment Videos
Influence of noise on wave boundary recognition by ECG measurement programs. Recommendations for preprocessing
J L Willems1, C Zywietz, P Arnaud
1University Hospital Gasthuisberg, Leuven, Belgium.
Computers and Biomedical Research, an International Journal
|December 1, 1987
Summary
This study tested how noise affects electrocardiogram (ECG) computer programs. Selective averaging is recommended for diagnostic ECG analysis, provided complex alignment is ensured.
Area of Science:
- Biomedical Engineering
- Medical Informatics
- Cardiology
Background:
- Quantitative Electrocardiography (QGE) analysis is crucial for cardiac diagnostics.
- Computerized ECG interpretation faces challenges from signal noise and variability.
- Standardized testing is needed to evaluate ECG algorithms under realistic conditions.
Purpose of the Study:
- To assess the impact of various noise types on ECG waveform detection by computer programs.
- To compare the performance of different ECG analysis programs under noisy conditions.
- To develop recommendations for robust ECG preprocessing and measurement strategies.
Main Methods:
- Systematic noise tests were conducted on 160 electrocardiograms (ECGs) using seven noise types.
- Eight ECG and six vectorcardiographic computer programs were evaluated.
- Program stability and precision were compared against noise-free recordings and expert cardiologist analysis.
Main Results:
- High-frequency noise caused shifts in P, QRS, and T wave detection by most programs.
- Programs utilizing selective averaging demonstrated significantly less variability in measurements.
- Averaging strategies were more stable than beat-to-beat or selected beat analysis.
Conclusions:
- Selective averaging is a recommended measurement strategy for diagnostic ECG computer programs.
- Proper alignment and waveform comparison are essential before averaging to ensure accuracy.
- These findings contribute to developing more reliable automated ECG analysis tools.