Applying machine learning techniques to electrocardiography and photoplethysmography signals and their multivariate-derived waveforms is an ongoing effort to estimate non-occlusive blood pressure. Unfortunately, real ambulatory electrocardiography and photoplethysmography waveforms are inevitably affected by motion and noise artifacts, so established machine learning architectures perform poorly when trained on data of the Multiparameter Intelligent Monitoring in Intensive Care II type, a publicly available ICU database. Our study addresses this problem by applying four well-established machine learning methods, i.e., random forest regression, support vector regression, Adaboost regression and artificial neural networks, to a small, self-sampled electrocardiography-photoplethysmography dataset (n = 54) to improve the robustness of machine learning to real-world BP estimates. We evaluated the performance using a selection of optimal feature morphologies of waveforms by using pulse arrival time, morphological and frequency photoplethysmography parameters and heart rate variability as characterization data. On the basis of the root mean square error and mean absolute error, our study showed that support vector regression gave the best performance for blood pressure estimation from noisy data, achieving an mean absolute error of 6.97 mmHg, which meets the level C criteria set by the British Hypertension Society. We demonstrate that ambulatory electrocardiography- photoplethysmography signals acquired by mobile discrete devices can be used to estimate blood pressure.
Citation: Mark Kei Fong Wong, Hao Hei, Si Zhou Lim, Eddie Yin-Kwee Ng. Applied machine learning for blood pressure estimation using a small, real-world electrocardiogram and photoplethysmogram dataset[J]. Mathematical Biosciences and Engineering, 2023, 20(1): 975-997. doi: 10.3934/mbe.2023045
Applying machine learning techniques to electrocardiography and photoplethysmography signals and their multivariate-derived waveforms is an ongoing effort to estimate non-occlusive blood pressure. Unfortunately, real ambulatory electrocardiography and photoplethysmography waveforms are inevitably affected by motion and noise artifacts, so established machine learning architectures perform poorly when trained on data of the Multiparameter Intelligent Monitoring in Intensive Care II type, a publicly available ICU database. Our study addresses this problem by applying four well-established machine learning methods, i.e., random forest regression, support vector regression, Adaboost regression and artificial neural networks, to a small, self-sampled electrocardiography-photoplethysmography dataset (n = 54) to improve the robustness of machine learning to real-world BP estimates. We evaluated the performance using a selection of optimal feature morphologies of waveforms by using pulse arrival time, morphological and frequency photoplethysmography parameters and heart rate variability as characterization data. On the basis of the root mean square error and mean absolute error, our study showed that support vector regression gave the best performance for blood pressure estimation from noisy data, achieving an mean absolute error of 6.97 mmHg, which meets the level C criteria set by the British Hypertension Society. We demonstrate that ambulatory electrocardiography- photoplethysmography signals acquired by mobile discrete devices can be used to estimate blood pressure.
| [1] |
P. M. Kearney, M. Whelton, K. Reynolds, P. Muntner, P. K Whelton, J. He, Global burden of hypertension: analysis of worldwide data, lancet, 365 (2005), 217–223. https://doi.org/10.1016/S0140-6736(05)17741-1 doi: 10.1016/S0140-6736(05)17741-1
|
| [2] |
G. A. Roth, M. H. Forouzanfar, A. E. Moran, R. Barber, G. Nguyen, V. L. Feigin, et al., Demographic and epidemiologic drivers of global cardiovascular mortality, N. Engl. J. Med., 372 (2015), 1333–1341. https://doi.org/10.1056/NEJMoa1406656 doi: 10.1056/NEJMoa1406656
|
| [3] |
K. T. Mills, A. Stefanescu, J. He, The global epidemiology of hypertension, Nat. Rev. Nephrol., 16 (2020), 223–237. https://doi.org/10.1038/s41581-019-0244-2 doi: 10.1038/s41581-019-0244-2
|
| [4] |
B. Williams, N. R. Poulter, M. J. Brown, M. Davis, G. T. McInnes, J. F. Potter, et al., British Hypertension Society guidelines for hypertension management 2004 (BHS-IV): Summary, BMJ, 328 (2004), 634–640. https://doi.org/10.1136/bmj.328.7440.634 doi: 10.1136/bmj.328.7440.634
|
| [5] | H. Y. Lee, T. Burkard, The advent of cuffless mobile device blood pressure measurement: Remaining challenges and pitfalls, Korean Circ. J., (2022), 52 (3): 198–204. https://doi.org/10.4070/kcj.2021.0405 |
| [6] |
R. Mieloszyk, H. Twede, J. Lester, J. Wander, S. Basu, G. Cohn, et al., A comparison of wearable tonometry, photoplethysmography, and electrocardiography for cuffless measurement of blood pressure in an ambulatory setting, IEEE J. Biomed. Health Inform., 26 (2022), 2864–2875. https://doi.org/10.1109/JBHI.2022.3153259 doi: 10.1109/JBHI.2022.3153259
|
| [7] | J. Esmaelpoor, Z. M. Sanat, M. H. Moradi, Blood Pressure Monitoring Using Photoplethysmogram and Electrocardiogram Signals, 1st edition, CRC Press, Florida, 2021. |
| [8] |
M. S. Tanveer, M. K. Hasan, Cuffless blood pressure estimation from electrocardiogram and photoplethysmogram using waveform based ANN-LSTM network, Biomed. Signal Proces., 51 (2019), 382–392. https://doi.org/10.1016/j.bspc.2019.02.028 doi: 10.1016/j.bspc.2019.02.028
|
| [9] |
M. Hosanee, G. Chan, K. Welykholowa, R. Cooper, P. A. Kyriacou, D. Zheng, et al., Cuffless single-site photoplethysmography for blood pressure monitoring, J. Clin. Med., 9 (2022), 723. https://doi.org/10.3390/jcm9030723 doi: 10.3390/jcm9030723
|
| [10] |
M. Elgendi, On the analysis of fingertip photoplethysmogram signals, Curr. Cardiol. Rev., 8 (2012), 14–25. https://doi.org/10.2174/157340312801215782 doi: 10.2174/157340312801215782
|
| [11] | E. Mejía-Mejía, J. Allen, K. Budidha, C. El-Hajj, P. A. Kyriacou, P. H. Charlton, 4- Photoplethysmography signal processing and synthesis, in Photoplethysmography, Academic Press, (2022), 69–146. https://doi.org/10.1016/B978-0-12-823374-0.00015-3 |
| [12] |
H. W. Loh, S. Xu, O. Faust, C. P. Ooi, P. D. Barua, S. Chakraborty, et al., Application of photoplethysmography signals for healthcare systems: An in-depth review, Comput. Methods Programs Biomed., 216 (2022), 106677. https://doi.org/10.1016/j.cmpb.2022.106677 doi: 10.1016/j.cmpb.2022.106677
|
| [13] |
R. C. Block, M. Yavarimanesh, K. Natarajan, A. Carek, A. Mousavi, A. Chandrasekhar, et al., Conventional pulse transit times as markers of blood pressure changes in humans, Sci. Rep., 10 (2020), 16373. https://doi.org/10.1038/s41598-020-73143-8 doi: 10.1038/s41598-020-73143-8
|
| [14] |
S. Heimark, O. M. H. Rindal, T. Seeberg, A. Stepanov, E. S. Boysen, C. L. Søraas, et al., Pulse arrival time can track changes in systolic blood pressure, J. Hypertens., 39 (2021), e132. https://doi.org/10.1097/01.hjh.0000745808.43316.c3 doi: 10.1097/01.hjh.0000745808.43316.c3
|
| [15] |
E. Finnegan, S. Davidson, M. Harford, J. Jorge, P. Watkinson, D. Young, et al., Pulse arrival time as a surrogate of blood pressure, Sci. Rep., 11 (2021), 22767. https://doi.org/10.1038/s41598-021-01358-4 doi: 10.1038/s41598-021-01358-4
|
| [16] | L. P. Yao, Z. l. Pan, Cuff-less blood pressure estimation from photoplethysmography signal and electrocardiogram, Phys. Eng. Sci. Med., 44 (2021), 397–408. https://doi.org/10.1007/s13246-021-00989-1 |
| [17] |
J, Lee, S. Yang, S. Lee, H. C. Kim, Analysis of pulse arrival time as an indicator of blood pressure in a large surgical biosignal database: Recommendations for developing ubiquitous blood pressure monitoring methods, J. Clin. Med., 8 (2019), 1773. https://doi.org/10.3390/jcm8111773 doi: 10.3390/jcm8111773
|
| [18] |
M. Puig-de-Morales-Marinkovic, K. T. Turner, J. P. Butler, J. J. Fredberg, S. Suresh, Viscoelasticity of the human red blood cell, Am. J. Physiol. Cell Physiol., 293 (2007), 597–605. https://doi.org/10.1152/ajpcell.00562.2006 doi: 10.1152/ajpcell.00562.2006
|
| [19] |
M. Nichelatti, P. Pettazzoni, G. Pallotti, The study of viscoelastic behavior of blood vessels, Blood Heart Circulation, 1 (2017), 1–3. https://doi.org/10.15761/BHC.1000110 doi: 10.15761/BHC.1000110
|
| [20] |
Z. Wang, M. J. Golob, N. Chesler, Viscoelastic properties of cardiovascular tissues, Viscoelastic Viscoplast. Mater., 2 (2016), 64. https://doi.org/10.5772/64169 doi: 10.5772/64169
|
| [21] |
R. Raghu, I. E. Vignon-Clementel, C. A. Figueroa, C. A. Taylor, Comparative study of viscoelastic arterial wall models in nonlinear one-dimensional finite element simulations of blood flow, J. Biomech. Eng., 133 (2011), 081003. https://doi.org/10.1115/1.4004532 doi: 10.1115/1.4004532
|
| [22] |
D. H. Bergel, The dynamic elastic properties of the arterial wall, J. Physiol., 156 (1961), 458–469. https://doi.org/10.1113/jphysiol.1961.sp006687 doi: 10.1113/jphysiol.1961.sp006687
|
| [23] | D. B. Camasã o, D. Mantovani, The mechanical characterization of blood vessels and their substitutes in the continuous quest for physiological-relevant performances. A critical review, Mater. Today Bio., 10 (2021), 100106. https://doi.org/10.1016/j.mtbio.2021.100106 |
| [24] |
S. Hodis, M. Zamir, Mechanical events within the arterial wall under the forces of pulsatile flow: A review, J. Mech. Behav. Biomed. Mater., 4 (2011), 1595–1602. https://doi.org/10.1016/j.jmbbm.2011.01.005 doi: 10.1016/j.jmbbm.2011.01.005
|
| [25] | S. L-O. Martin, A. M. Carek, C-S. Kim, H. Ashouri, O. T. Inan, J-O Hahn, et al., Weighing scale-based pulse transit time is a superior marker of blood pressure than conventional pulse arrival time. Sci. Rep., 8 (2018), 15838. https://doi.org/10.1038/srep39273 |
| [26] | R. Barbieri, E. P. Scilingo, G. Valenza, Complexity and nonlinearity in cardiovascular signals, Springer, Berlin, 2017. |
| [27] | S. G. Khalid, J. Zhang, F. Chen, D. Zheng, Blood pressure estimation using photoplethysmography only: comparison between different machine learning approaches. J. Healthc. Eng., (2018), 1548647. https://doi.org/10.1155/2018/1548647 |
| [28] |
M. H. Chowdhury, M. N. I. Shuzan, M. E. H. Chowdhury, Z. B. Mahbub, M. M. Uddin, A. Khandakar, et al., Estimating blood pressure from the photoplethysmogram signal and demographic features using machine learning techniques, Sensors, 20 (2020), 3127. https://doi.org/10.3390/s20113127 doi: 10.3390/s20113127
|
| [29] |
A. Chakraborty, D. Sadhukhan, S. Pal, M. Mitra, PPG-based automated estimation of blood pressure using patient-specific neural network modeling, J. Mech. Med. Biol., 20 (2020), 2050037. https://doi.org/10.1142/S0219519420500372 doi: 10.1142/S0219519420500372
|
| [30] |
A. S. Zadi, R. Alex, R. Zhang, D. E. Watenpaugh, K. Behbehani, Arterial blood pressure feature estimation using photoplethysmography, Comput. Biol. Med., 102 (2018), 104–111. https://doi.org/10.1016/j.compbiomed.2018.09.013 doi: 10.1016/j.compbiomed.2018.09.013
|
| [31] |
Z. Liu, B. Zhou, Y. Li, M. Tang, F. Miao, Continuous blood pressure estimation from electrocardiogram and photoplethysmogram during arrhythmias, Front. Physiol., 11 (2020), 575407. https://doi.org/10.3389/fphys.2020.575407 doi: 10.3389/fphys.2020.575407
|
| [32] | C. Sideris, H. Kalantarian, E. Nemati, M. Sarrafzadeh, Building continuous arterial blood pressure prediction models using recurrent networks, in 2016 IEEE International Conference on Smart Computing (SMARTCOMP), (2016), 1–5. https://doi.org/10.1109/SMARTCOMP.2016.7501681 |
| [33] |
H. Eom, D. Lee, S. Han, Y. S. Hariyani, Y. Lim, I. Sohn, et al., End-to-end deep learning architecture for continuous blood pressure estimation using attention mechanism, Sensors, 20 (2020), 2338. https://doi.org/10.3390/s20082338 doi: 10.3390/s20082338
|
| [34] |
A. Paviglianiti, V. Randazzo, S. Villata, G. Cirrincione, E. Pasero, A comparison of deep learning techniques for arterial blood pressure prediction, Cognit. Comput., 14 (2021), 1689–1710. https://doi.org/10.1007/s12559-021-09910-0 doi: 10.1007/s12559-021-09910-0
|
| [35] |
Y. Li, L. N. Harfiya, K. Purwandari, Y. Lin, Real-time cuffless continuous blood pressure estimation using deep learning model, Sensors, 20 (2020), 5606. https://doi.org/10.3390/s20195606 doi: 10.3390/s20195606
|
| [36] |
C. El-Hajj, P. A. Kyriacou, A review of machine learning techniques in photoplethysmography for the non-invasive cuff-less measurement of blood pressure, Biomed. Signal Proces., 58 (2020), 101870. https://doi.org/10.1016/j.bspc.2020.101870 doi: 10.1016/j.bspc.2020.101870
|
| [37] |
A. Papaa, M. Mitalb, P. Pisanoa, M. D. Giudice, E-health and wellbeing monitoring using smart healthcare devices: An empirical investigation, Technol. Forecast. Soc., 153 (2020), 119226. https://doi.org/10.1016/j.techfore.2018.02.018 doi: 10.1016/j.techfore.2018.02.018
|
| [38] |
S. H. Chuah, P. A. Rauschnabel, N. Krey, B. Nguyen, T. Ramayah, S. Lade, Wearable technologies: The role of usefulness and visibility in smartwatch adoption, Compu. Hum. Behav., 65 (2016), 276–284. https://doi.org/10.1016/j.chb.2016.07.047 doi: 10.1016/j.chb.2016.07.047
|
| [39] |
M. Saeed, M. Villarroel, A. T Reisner, G. Clifford, L. Lehman, G. Moody, et al., Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II): A public-access intensive care unit database, Crit. Care Med., 39 (2011) 952–960. https://doi.org/10.1097/CCM.0b013e31820a92c6 doi: 10.1097/CCM.0b013e31820a92c6
|
| [40] | A. L. Goldberger, L. A. Amaral, L. Glass, J. M. Hausdorff, P. C. Ivanov, R. G. Mark, et al., PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals, circulation, 101 (2000), 215–220. https://doi.org/10.1161/01.cir.101.23.e215 |
| [41] |
E. Martinez-Ríosa, L. Montesinosa, M. Alfaro-Poncea, L. Pecchia, A review of machine learning in hypertension detection and blood pressure estimation based on clinical and physiological data, Biomed. Signal Proces., 68 (2021), 102813. https://doi.org/10.1016/j.bspc.2021.102813 doi: 10.1016/j.bspc.2021.102813
|
| [42] |
D. U. Jeong, K. M. Lim, Combined deep CNN–LSTM network-based multitasking learning architecture for noninvasive continuous blood pressure estimation using difference in ECG-PPG features, Sci. Rep., 11 (2021), 13539. https://doi.org/10.1038/s41598-021-92997-0 doi: 10.1038/s41598-021-92997-0
|
| [43] | Ü. Şentürk, I. Yücedağ, K. Polat, Repetitive neural network (RNN) based blood pressure estimation using PPG and ECG signals, in 2018 2nd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), (2018), 1–4. https://doi.org/10.1109/ISMSIT.2018.8567071 |
| [44] |
I. Eşer, L. Khorshid, U. Y. Güneş, Y. Demir, The effect of different body positions on blood pressure, J. Clin. Nurs., 16 (2007), 137–140. https://doi.org/10.1111/j.1365-2702.2005.01494.x doi: 10.1111/j.1365-2702.2005.01494.x
|
| [45] |
J. A. Sukor, S. J. Redmond, N. H. Lovell, Signal quality measures for pulse oximetry through waveform morphology analysis, Physiol. Meas., 32 (2011), 369–384. https://doi.org/10.1088/0967-3334/32/3/008 doi: 10.1088/0967-3334/32/3/008
|
| [46] | Paul van, HeartPy-Python Heart Rate Analysis Toolkit, https://python-heart-rate-analysis-toolkit.readthedocs.io/en/latest/ (11-July-2022) |
| [47] | J. Cano, A. Quesada, F. Ravelli, R. Zangróniz, R. Alcaraz, J. J. Rieta, Novel photoplethysmographic and electrocardiographic features for enhanced detection of hypertensive individuals, in 2021 International Conference on e-Health and Bioengineering (EHB), (2021), 1–4. https://doi.org/10.1109/EHB52898.2021.9657546 |
| [48] |
S. Chen, Z. Ji, H. Wu, Y. Xu, A non-invasive continuous blood pressure estimation approach based on machine learning, Sensors (Basel), 19 (2019), 2585. https://doi.org/10.3390/s19112585 doi: 10.3390/s19112585
|
| [49] | T. T. Nguyen, J. Z. Huang, T. T. Nguyen, Unbiased feature selection in learning random forests for high-dimensional data, Sci. World J., (2015), e471371. https://doi.org/10.1155/2015/471371 |
| [50] |
M. Elgendi, Y. Liang, R. Ward, Toward generating more diagnostic features from photoplethysmogram waveforms, Diseases, 6 (2018), 20. https://doi.org/10.3390/diseases6010020 doi: 10.3390/diseases6010020
|
| [51] | J. Dey, A. Gaurav, V. N. Tiwari, InstaBP: Cuff-less blood pressure monitoring on smartphone using single PPG sensor, in 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), (2018), 5002–5005. https://doi.org/10.1109/embc.2018.8513189 |
| [52] | H. Tjahjadi, K. Ramli, H. Murfi, Noninvasive classification of blood pressure based on photoplethysmography signals using bidirectional long short-term memory and time-frequency analysis, IEEE Access, 8 (2020). https://doi.org/10.1109/ACCESS.2020.2968967 |
| [53] |
T. Vandenberk, J. Stans, C. Mortelmans, R. Van Haelst, G. V. Schelvergem, C. Pelckmans, et al., Clinical validation of heart rate apps: mixed-methods evaluation study, JMIR Mhealth Uhealth, 5 (2017), e129. https://doi.org/10.2196/mhealth.7254 doi: 10.2196/mhealth.7254
|
| [54] | P. A. Lanfranchi, V. K. Somers, Cardiovascular physiology: autonomic control in health and in sleep disorders, in Principles and Practice of Sleep Medicine, Elsevier, (2017), 142–154. https://doi.org/10.1016/B978-0-323-24288-2.00014-3 |
| [55] |
N. Selvaraj, A. Jaryal, J. Santhosh, K. K. Deepak, S. Anand, Assessment of heart rate variability derived from finger-tip photoplethysmography as compared to electrocardiography, J. Med. Eng. Tech., 32 (2008), 479–484. https://doi.org/10.1080/03091900701781317 doi: 10.1080/03091900701781317
|
| [56] | A. Tiloca, G. Pagana, D. Demarchi, A random tree based algorithm for blood pressure estimation, in 2020 IEEE MTT-S International Microwave Biomedical Conference (IMBioC), (2020), 1–4. https://doi.org/10.1109/IMBIoC47321.2020.9385038 |
| [57] | Y. Zhang, Z. Feng, A SVM method for continuous blood pressure estimation from a PPG signal, in Proceedings of the 9th International Conference on Machine Learning and Computing, (2017), 128–132. https://doi.org/10.1145/3055635.3056634 |
| [58] |
M. Kachuee, M. M. Kiani, H. Mohammadzade, M. Shabany, Cuffless blood pressure estimation algorithms for continuous health-care monitoring, IEEE Trans. Biomed. Eng., 64 (2017), 859–869. https://doi.org/10.1109/TBME.2016.2580904 doi: 10.1109/TBME.2016.2580904
|
| [59] |
E. O'Brien, J. Petrie, W. Littler, M. de Swiet, P. L. Padfield, K. O'Malley, et al., The british hypertension Society protocol for the evaluation of automated and semi-automated blood pressure measuring devices with special reference to ambulatory systems, J. Hypertens., 8 (1990), 607–619. https://doi.org/10.1097/00004872-199007000-00004 doi: 10.1097/00004872-199007000-00004
|
| [60] |
E. O'Brien, B. Waeber, G. Parati, J. Staessen, M. G. Myers, Blood pressure measuring devices: Recommendations of the European society of hypertension, BMJ, 322 (2001), 531–536. https://doi.org/10.1136/bmj.322.7285.531 doi: 10.1136/bmj.322.7285.531
|
| [61] | L. Wang, W. Zhou, Y. Xing, X. Zhou, A novel neural network model for blood pressure estimation using photoplethesmography without electrocardiogram, J. Healthc. Eng., (2018), 1–9. https://doi.org/10.1155/2018/7804243 |
| [62] |
Q. Yousef, M. B. I. Reaz, M. A. M. Ali, The analysis of PPG morphology: investigating the effects of aging on arterial compliance, Meas. Sci. Rev., 12 (2012), 266–271. https://doi.org/10.2478/v10048-012-0036-3 doi: 10.2478/v10048-012-0036-3
|
Figures(14) / Tables(5)
Mark Kei Fong Wong, Hao Hei, Si Zhou Lim, Eddie Yin-Kwee Ng. Applied machine learning for blood pressure estimation using a small, real-world electrocardiogram and photoplethysmogram dataset[J]. Mathematical Biosciences and Engineering, 2023, 20(1): 975-997. doi: 10.3934/mbe.2023045
DownLoad: