Removal of cardiac beat artifact in oesophageal pressure measurement by frequency analysis

Abstract

Oesophageal pressure (Pes) measurements are important in medical research and useful in clinical diagnosis. Measurements, however, are contaminated heavily by cardiac artifacts. The spectrum and waveform of the Pes signal is obtained from the oesophageal balloon. Adaptive finite impulse response (AFIR) filter and modified adaptive noise cancellation (MANC) methods are adopted to filter out cardiac beat interference. These results are compared. In the frequency domain, frequency variations and spectral overlap between the Pes components and cardiac beat signal components impact on the performance of the filter. From our experimental results on power strength, the fourth or higher harmonics did not have any significant effect on the filter performance. However, the second harmonics of these signals had a significant effect on the filtering result. Thus, in the design of AFIR filters, attention is needed to minimise these effects. In frequency analysis, these harmonics or overlapping frequencies do not affect MANC. MANC was the better method for eliminating cardiac beat artifact in Pes measurement. The dynamic compliance (Cdyn) was also used to evaluate the performance of MANC and AFIR. The standard deviation of Cdyn was less than 0.15 using MANC, compared with standard deviations as high as 0.57 for AFIR. We conclude that MANC performs better than AFIR.

A Singular Spectrum Analysis-Based Data-Driven Technique for the Removal of Cardiogenic Oscillations in Esophageal Pressure Signals

“…A direct comparison of performance of the proposed signal denoising technique with that of other techniques which are reported in [4], [6], [7], [8] and [9] is not possible as the performance these techniques are evaluated with different settings and databases. However, in [6], the simulated signals were added with artificial and white Gaussian noise (WGN) to have a signal-to-noise ratio of 10 dB, and then the noisy signals are denoised.…”

“…A direct comparison of performance of the proposed \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${ \boldsymbol {P}}_{ \boldsymbol {eso}}$ \end{document} signal denoising technique with that of other techniques which are reported in [4] and [6] – [9] is not possible as the performance these techniques are evaluated with different settings and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${ \boldsymbol {P}}_{ \boldsymbol {eso}}$ \end{document} databases. However, in [6] , the simulated \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${ \boldsymbol {P}}_{ \boldsymbol {eso}}$ \end{document} signals were added with artificial \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$\boldsymbol {CGO}$ \end{document} and white Gaussian noise (WGN) to have a signal-to-noise ratio of 10 dB, and then the noisy signals are denoised.…”

“…Therefore, it is difficult to premise the performance of the MANC technique on signals collected from intensive care unit patients. An enhanced version of the MANC technique is reported in [8].…”

“…Therefore, it is difficult to premise the performance of the MANC technique on \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${ \boldsymbol {P}}_{ \boldsymbol {eso}}$ \end{document} signals collected from intensive care unit patients. An enhanced version of the MANC technique is reported in [8] .…”

“…All the above discussed techniques [4] , [6] – [9] (except [10] ), require an additional signal, be it an ECG or EMG, along with \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}${ \boldsymbol {P}}_{ \boldsymbol {eso}}$ \end{document} . Acquisition of an extra signal undoubtedly increases the complexity of the system and hinders patients’ comfort.…”