Journal of Biomedical Engineering Research (대한의용생체공학회:의공학회지)

The Korean Society of Medical and Biological Engineering (대한의용생체공학회)
권호 표지

Design of Neural Network Based IEF Filter for Time-varying Control of Incremental Factor

증가인자 시변제어를 위한 신경망 증가평가필터 설계

  • 박상희 (금오공과대학교 전자공학부) ;
  • 최한고 (금오공과대학교 전자공학부)
  • Published : 2002.11.01
Download PDF
⟨ Previous Next ⟩

Abstract

Powerline interference in bioelectric recordings is a common source of noise. IEF(Incremental Estimation Filter) has been used to eliminate powerline interferences in biosignals, especially in ECG(Electrocadiogram) signals. The constant incremental factor in the IEF filter, which affects the performance of noise rejection, is usually determined empirically or experimentally based on the input signals. This paper presents the design of neural network based IEF filter for time-varying control of the incremental factor. The proposed IEF filter is evaluated by applying to artificial signals as well as ECG signals of MIT-BIH database. For the relative comparison of noise-rejection performance, it is compared with adaptive noise canceler and conventional IEF filter. Simulation results show that the neural network based IEF filter outperforms these adaptive filters with respect to convergence speed and noise rejection is specific frequencies.

생체신호 수집시 전력선 잡음은 일반적인 잡음원이다. 증가평가필터(Incremental Estimation Filter. IEF)는 생체신호, 특히, 심전도 (Electrocadiogram, ECG) 신호에 있어서 전력선 잡음을 제거하기 위해 사용되어 왔다. 증가평가필터의 잡음제거 성능에 영향을 미치는 상수 값의 증가인자는 입력신호에 따라서 경험적으로 혹은 실험적으로 결정되고 있다. 본 논문에서는 증가인자의 시변(time-varying) 제어를 위해 신경망을 이용한 증가평가필터 설계를 제시하고 있다. 제안된 증가평가필터는 인위적인 신호뿐만 아니라 MIT/BIH 데이터베이스의 실제 심전도 신호에 적용함으로써 평가하였으며, 잡음제거 성능의 상대적인 비교를 위해 적응잡음제거기와 기존의 증가평가필터등과 비교하였다. 실험결과 신경망에 근거한 증가평가필터는 수렴속도와 특정 주파수에서의 잡음제거에서 기존의 적응필터보다 우수함을 보여주었다.

Keywords

References

  1. IEEE Trans. Biomed. Eng. v.20 60Hz interference in electrocardiography J.C. Huhta;J. G. Webster https://doi.org/10.1109/TBME.1973.324169
  2. Upper Saddle River Discrete Time Signal Processing A.V. Oppenheim;R. W. Schafer
  3. Med. Biol. Eng v.9 Recursive digital filter for biological signals P.A. Lynn https://doi.org/10.1007/BF02474403
  4. IEEE Trans. Audio Electroacoust v.16 no.3 Ditital filtering with applications to electrocardiogram processing C.S. Weaver;R.W. Lawrence
  5. IEEE Trans. Biomed. Eng v.32 no.9 Digital filters for real-time ECG signal processing using microprocessors M.L. Alstorm;w.J. Tompkins https://doi.org/10.1109/TBME.1985.325589
  6. IEEE Trans. Biomed. Eng v.43 A comparison of adaptive and nonadaptive filters for reductions of power line interference in the ECG P.S. Hamilton https://doi.org/10.1109/10.477707
  7. KOSOMBE v.17 no.1 A study on the elimination of 60Hz powerline interference for the automated diagnosis of electrocadiogram H.J. Kweon;K.S. Jeong;M.H. Lee
  8. Computer in Cardiology Digital filters for real-time ECG signals D.W. Mortara
  9. Adaptive signal Proessing Adaptive signal Proessing B. Widrow;S.D. Stearns
  10. Math. Infrom. Med. v.33 no.1 Interference removal with an improved incremental estimation filter J.A. Kors
  11. Biomedical Digital Signal Processing Biomedical Digital Signal Processing W.J. Tomkins
  12. IEEE Trans. Biomed. Eng. v.37 A new technique for line interference monitoring and reduction in biopotential amplifiers Y.Z. Ider;H. Koymen https://doi.org/10.1109/10.55666
  13. IEEE Eng. Med. and Bio v.1 Physionet:A web-based resource for the study physiologic signals G.B. Moody;R.G. Mark and A.L Glodberger
  14. Design of Microcomputer-Based Medical Instrumentation Design of Microcomputer-Based Medical Instrumentation W.J. Tomkins;J.G. Webster