Estimating Blood Pressure with a Smartphone

  • Rong-Chao Peng
  • Wen-Rong Yan
  • Xiao-Lin Zhou
Conference paper
Part of the IFMBE Proceedings book series (IFMBE, volume 64)


Cardiovascular disease like hypertension is one of the top killers of human’s life.

1 Introduction

Cardiovascular disease like hypertension is one of the top killers of human’s life. As the smartphone is becoming ubiquitous in the world, its application in medicine may provide an easy and low-cost approach for early detection of cardiovascular disease. Lamonaca et al. applied artificial neural network to evaluate blood pressure from the pulse wave signal acquired by the camera of the smartphone [1]. Chandrasekaran et al. proposed two vascular transit time methods to cufflessly estimate blood pressure with smartphones [2]. In this study, we proposed a novel method to estimate continuous blood pressure from heart sound signals acquired by the built-in microphone of the smartphone.

2 Methods

It is known in clinical medicine that the second heart sound (S2) has a characteristic “accentuation” in hypertensive patients. We thus infer that the pattern of the S2 is associated with the blood pressure. If the pattern of the S2 is recognized, then the blood pressure might be determined.

Thirty-two healthy subjects participated in the experiment (25 males and 7 females, age 20–32 years, height 150–185 cm, weight 44–90 kg). They were instructed to lie in the supine position on a mattress. A smartphone was used to collect the heart sound signals and simultaneously a Finometer® MIDI (Model II, Finapres Medical Systems B.V., The Netherlands) was used to measure continuous beat-to-beat blood pressure.

Each heart sound signal was down-sampled to 2205 Hz and its envelope of Shannon energy was calculated. The peaks of the envelope were identified by a high threshold and a low threshold, and then were classified as the first heart sound (S1) or S2 according to the clinical knowledge that the duration from S1 to S2 is shorter than that from S2 to S1.

Afterwards, the S2 of each heartbeat in the heart sound signals was truncated by a 64 ms window centered at the maximum of the S2, and its frequency spectrum was obtained by Fast Fourier transform (FFT). The spectrum was then normalized by dividing its maximum value, and the 36 spectral values with 10 Hz interval in the frequency band 50–400 Hz were chosen as features of the S2. These features were separately regressed with the systolic and diastolic blood pressure using support vector machine (SVM) developed by Chang and Lin [3]. The accuracy of the regression was evaluated using tenfold cross-validation.

3 Results

The predicted values of the systolic and diastolic blood pressure using the SVM regression were compared with the corresponding values measured by the Finometer device. Figure 1 shows that the correlation coefficients between the predicted values and the measured values were 0.893 and 0.922 for systolic and diastolic blood pressure, respectively. These results suggest that the smartphone is of potential use for cuffless and continuous blood pressure monitoring.
Fig. 1

Correlation analysis between predicted values and measured values for (a) systolic blood pressure (SBP) and (b) diastolic blood pressure (DBP). In each panel, the diagonal is the linear regression line. r, Pearson correlation coefficient

4 Conclusion

We demonstrated that the smartphone can be used to estimate blood pressure from the heart sound signals, continuously, non-invasively and cufflessly.



This work was supported in part by the National Natural Science Foundation of China (no. 61401453), the STS Key Health Program of Chinese Academy of Sciences (no. KFJEW-STS-097 and KFJ-EW-STS-095), the External Cooperation Program of the Chinese Academy of Sciences (GJHZ1212), the Guangdong Innovation Research Team Fund for Low-Cost Healthcare Technologies in China, the Key Lab for Health Informatics of Chinese Academy of Sciences, and the Enhancing Program of Key Laboratories of Shenzhen City (no. ZDSY20120617113021359).


  1. 1.
    F. Lamonaca, K. Barbe, Y. Kurylyak, et al., “Application of the Artificial Neural Network for blood pressure evaluation with smartphones,” in 2013 IEEE 7th Intl. Conf. on Intell. Data Acq. and Advanced Comput. Sys., IDAACS 2013, Sep 12, 2013 - Sep 14, 2013, Berlin, Germany, 2013, pp. 408–412.Google Scholar
  2. 2.
    V. Chandrasekaran, R. Dantu, S. Jonnada, et al, “Cuffless differential blood pressure estimation using smart phones,” IEEE Trans. on Biomed. Eng., vol. 60, pp. 1080–1089, Apr 2013.CrossRefGoogle Scholar
  3. 3.
    C.-C. Chang and C.-J. Lin, “LIBSVM: A library for support vector machines,” Acm Trans. on Intell. Sys. and Tech., vol. 2, pp. 1–27, 2011.CrossRefGoogle Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  1. 1.Chinese Academy of SciencesShenzhen Institutes of Advanced TechnologyShenzhenChina

Personalised recommendations