Publication detail

Adaptive recurrent system for noise cancellation and arrhythmia detection

PROVAZNÍK, I. HOLČÍK, J.

Original Title

Adaptive recurrent system for noise cancellation and arrhythmia detection

Type

conference paper

Language

English

Original Abstract

A new LMS-based adaptive filter for preprocessing and analysis of ECG signals recorded under conditions of stress-tests is introduced. The adaptive filter was developed to cancel motion artifacts and myopotentials and to detect diagnostically significant waves in the ECG signal, e.g. arrhythmias. The new filter exploits a repetitivity of the ECG signal for optimal removing of the nonstationary noise. It uses a vector-variable-step-size LMS algorithm to adjust filter weights according to the nonstationary properties of the processed signal. The developed system was tested on real signals and on the CSE signal library, and the results are described.

Keywords

adaptive filter, ECG signal, arrhythmias

Authors

PROVAZNÍK, I.; HOLČÍK, J.

RIV year

2004

Released

1. 10. 1994

Publisher

IEEE

Location

Baltimore

Pages from

1270

Pages to

1271

Pages count

2

BibTex

@inproceedings{BUT11303,
  author="Valentine {Provazník} and Jiří {Holčík}",
  title="Adaptive recurrent system for noise cancellation and arrhythmia detection",
  booktitle="Proceedings of the 16th Annual International Conference of the IEEE EMBS",
  year="1994",
  pages="1270--1271",
  publisher="IEEE",
  address="Baltimore"
}