Automatic assessment of blood pressure for Korotkoff sounds on the basis of human hearing threshold

Abstract

Objective In this study, to measure blood pressure (BP) on the basis of human hearing threshold, we proposed a method that detects the audible or inaudible Korotkoff sounds (K-sounds) using the equal loudness contour and automatically assesses the BP.

Methods In this study, we detected the systolic period of K-sounds using cuff pressure oscillation and then converted the K-sounds corresponding to the systolic interval into sound pressure levels (SPLs). Next, the systolic blood pressure (SBP) and diastolic blood pressure (DBP) were assessed by mapping the K-sounds, which were converted into SPLs on an equal loudness contour.

Results To validate the accuracy of our proposed method, we compared it with the auscultatory method. The mean differences (mean±SD) in the SBP and DBP were 0.31±1.95 and 1.20±2.17 mmHg, respectively. For the SBP, the linear regression equation was y=0.98x+1.56 mmHg (where x and y represent the auscultatory and the proposed method, respectively), with a SE of estimate of 1.93 mmHg and a correlation coefficient of 0.99. For the DBP, the linear regression equation was y=1.01x−1.94 mmHg, with an SE of estimate of 2.18 mmHg and a correlation coefficient of 0.98. All P values were less than 0.0001 for both regressions.

Conclusion The auscultatory method of BP monitoring is sensitive to the observer’s condition or environmental noise. To overcome these disadvantages, we used the human hearing threshold for objective SBP and DBP automatic assessment, and this method can be applicable to an automatic auscultatory method.