Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Nov 8;22(22):8607.
doi: 10.3390/s22228607.

Machine-Learning Classification of Pulse Waveform Quality

Affiliations

Machine-Learning Classification of Pulse Waveform Quality

Te Ouyoung et al. Sensors (Basel). .

Abstract

Pulse measurements made using wearable devices can aid the monitoring of human physiological condition. Accurate estimation of waveforms is often difficult for nonexperts; motion artifacts may occur during tonometry measurements when the skin-sensor contact pressure is insufficient. An alternative approach is to extract only high-quality pulses for use in index calculations. The present study aimed to determine the effectiveness of using machine-learning analysis in discriminating between high-quality and low-quality pulse waveforms induced by applying different contact pressures. Radial blood pressure waveform (BPW) signals were measured noninvasively in healthy young subjects using a strain-gauge transducer. One-minute-long trains of pulse data were measured when applying the appropriate contact pressure (67.80 ± 1.55 mmHg) and a higher contact pressure (151.80 ± 3.19 mmHg). Eight machine-learning algorithms were employed to evaluate the following 40 harmonic pulse indices: amplitude proportions and their coefficients of variation and phase angles and their standard deviations. Significant differences were noted in BPW indices between applying appropriate and higher skin-surface contact pressures. The present appropriate contact pressure could not only provide a suitable holding force for the wearable device but also helped to maintain the physiological stability of the underlying tissues. Machine-learning analysis provides an effective method for distinguishing between the high-quality and low-quality pulses with excellent discrimination performance (leave-one-subject-out test: random-forest AUC = 0.96). This approach will aid the development of an automatic screening method for waveform quality and thereby improve the noninvasive acquisition reliability. Other possible interfering factors in practical applications can also be systematically studied using a similar procedure.

Keywords: contacting pressure; machine learning; pulse; spectral analysis; waveform quality.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
Typical measured pulse waveforms for (a) Data A (high quality) and (b) Data B (low quality).
Figure 2
Figure 2
Flowchart of information processing in the study. Forty indices (Cn, Pn, CVn, and Pn_SD; n = 1–10) were calculated, which were used as the features for the information processing. Threefold cross-validation was used to evaluate the model performance, and then leave-one-subject-out analysis was used to test the model accuracy.
Figure 3
Figure 3
Comparisons of harmonic indices between applying the appropriate and higher contact pressures: Cn, CVn, Pn, and Pn_SD. Data are mean and standard-deviation values. C6C10 values have been multiplied by 10 to make the differences clearer. “*” indicates p < 0.05; “+” indicates 0.05 < p < 0.1. All Cn and CV1CV9 were significantly larger when applying higher-than-appropriate contact pressure. Regarding phase-angle indices, there were significant differences in P2, P4, and P6P10. Regarding Pn_SD indices, P2_SDP10_SD were either significantly (p < 0.05) or marginally (0.05 < p < 0.1) larger when applying a higher-than-appropriate contact pressure.
Figure 3
Figure 3
Comparisons of harmonic indices between applying the appropriate and higher contact pressures: Cn, CVn, Pn, and Pn_SD. Data are mean and standard-deviation values. C6C10 values have been multiplied by 10 to make the differences clearer. “*” indicates p < 0.05; “+” indicates 0.05 < p < 0.1. All Cn and CV1CV9 were significantly larger when applying higher-than-appropriate contact pressure. Regarding phase-angle indices, there were significant differences in P2, P4, and P6P10. Regarding Pn_SD indices, P2_SDP10_SD were either significantly (p < 0.05) or marginally (0.05 < p < 0.1) larger when applying a higher-than-appropriate contact pressure.
Figure 4
Figure 4
Comparisons of harmonic indices between Data A (high quality) and Data B (low quality): Cn, CVn, Pn, and Pn_SD. Data are mean and standard-deviation values. C6C10 values have been multiplied by 10 to make the differences clearer. “*” indicates p < 0.05. “+” indicates 0.05 < p < 0.1. Cn, CVn, and Pn_SD were all significantly larger for Data B than for Data A. Regarding phase-angle indices, there were significant differences in P2, P5, and P7P10.

Similar articles

References

    1. Dresher R.P., Mendelson Y. Reflectance forehead pulse oximetry: Effects of contact pressure during walking. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2006;2006:3529–3532. - PubMed
    1. Takeshita T., Yoshida M., Takei Y., Ouchi A., Hinoki A., Uchida H., Kobayashi T. Relationship between Contact Pressure and Motion Artifacts in ECG Measurement with Electrostatic Flocked Electrodes Fabricated on Textile. Sci. Rep. 2019;9:5897. doi: 10.1038/s41598-019-42027-x. - DOI - PMC - PubMed
    1. Wilkinson I.B., Cockcroft J.R., Webb D.J. Pulse wave analysis and arterial stiffness. J. Cardiovasc. Pharmacol. 1998;32((Suppl. 3)):S33–S37. - PubMed
    1. O’Rourke M.F., Adji A., Safar M.E. Structure and Function of Systemic Arteries: Reflections on the Arterial Pulse. Am. J. Hypertens. 2018;31:934–940. doi: 10.1093/ajh/hpy084. - DOI - PubMed
    1. Johnson J.E., Shay O., Kim C., Liao C. Wearable Millimeter-Wave Device for Contactless Measurement of Arterial Pulses. IEEE Trans. Biomed. Circuits Syst. 2019;13:1525–1534. doi: 10.1109/TBCAS.2019.2948581. - DOI - PubMed

LinkOut - more resources