Related Experiment Videos
Detection and deletion of motion artifacts in electrogastrogram using feature analysis and neural networks
J Liang1, J Y Cheung, J D Chen
1Institute for Healthcare Research, Baptist Medical Center, Oklahoma City 73112, USA.
Annals of Biomedical Engineering
|September 23, 1997
Summary
This study introduces an automated method using neural networks to detect and remove motion artifacts from electrogastrogram (EGG) recordings. This noninvasive technique improves the analysis of gastric myoelectrical activity.
Area of Science:
- Biomedical Engineering
- Physiological Measurement
- Signal Processing
Background:
- Electrogastrography (EGG) is a noninvasive method to study gastric electrical activity.
- Motion artifacts significantly degrade EGG signal quality, hindering analysis.
- Current artifact removal relies on time-consuming and subjective visual inspection.
Purpose of the Study:
- To develop an automated method for detecting and eliminating motion artifacts in EGG signals.
- To improve the efficiency and objectivity of EGG data analysis.
- To enable reliable physiological and pathophysiological studies of the stomach.
Main Methods:
- Feature analysis was employed to identify characteristics of motion artifacts.
- Artificial neural networks were utilized for automated artifact detection.
- Different feature combinations were evaluated to optimize neural network performance.
Main Results:
- A novel computer-based method for automatic motion artifact detection in EGG was developed.
- The study investigated and characterized various types of motion artifacts.
- Optimal feature sets for neural network input were identified for improved detection accuracy.
Conclusions:
- The developed automated method offers an objective and efficient alternative to manual artifact removal.
- This technique enhances the reliability of electrogastrography for clinical and research applications.
- The noninvasive study of gastric function can be significantly advanced by this artifact elimination approach.