Adaptable Noise Reduction of ECG Signals for Feature Extraction
Conventionally, ECG signal is measured on static condition since various types of noise including muscle artifact noise and electrode moving artifact noise are in coupled in dynamic environment. To solve this problem, various noised signals are grouped into six categories by context estimation, and effectively reconfigured noise reduction filter by neural network and genetic algorithm (GA). Neural network based control module effectively select optimal filter block by noise context based clustering at running mode and filtering performance was improved by GA at evolution mode. Experimental results showed that proposed algorithm effectively remove baseline wander noise and muscle noise and feature extraction results showed significant improvement of T duration extraction values.
KeywordsFinite Impulse Response Decision Module Impulsive Noise Artifact Noise Noise Reduction Algorithm
Unable to display preview. Download preview PDF.
- 1.Park, K., Lee, K., Lee, B., Jeong, K., Yoon, H.: Design of A Wavelet Interpolation Filter for Elimination of Muscle Artifact in the Stress ECG. J. Biomedical Eng. Res. 21(5), 495–503 (2000)Google Scholar
- 2.Yoon, S., Park, H., Park, J., Bae, E., Kim, Y.: An Implementation of Digital Signal Process-ing System for The Baseline Elimination. J. KIMICS 5(7), 1287–1294 (2001)Google Scholar
- 3.Pancer, T.P.: A Suppression of an Impulsive Noise in ECG Signal Processing. In: Proc. 26th Annual Int’l. Conf. IEEE EMBS, pp. 596–599 (2004)Google Scholar
- 4.Kim, H., Min, C., Kim, T.: Feature Extraction of ECG Signal for Heart Disease Diagnosis. In: Proc. IEEK & KIEE Conf. Information and Control Systems, pp. 325–327 (2004)Google Scholar