Cuffless Blood Pressure Monitoring: Experimental Evidences of a Beat-to-Beat PPG Technique

Abstract

While elevated blood pressure affects one third of adults in developed countries, there is nowadays no comfortable and easy-to-deploy technology to assess blood pressure in daily life scenarios. The increasing use of smart devices has recently spread the availability of photoplethysmographic (PPG) sensors at everyone's wrist (smartwatches), or eventually the finger (camera-based sensing on smartphones). Although such PPG sensors represent a strong opportunity to assess physiological information on mobile device users, no reliable technique has so far been demonstrated to exploit such PPG signals for continuous cuffless blood pressure monitoring.
We present the results of a novel technique that provides cuffless beat-to-beat estimates of blood pressure from PPG signals with high level of accuracy. This optical technique relies on the physiological analysis of PPG signals acquired either in transmission or reflection mode. The developed algorithm extracts features from PPG pulses based on the wave reflection theory of the arterial system. The calculated features are further converted into blood pressure values via a calibration procedure.
Performance results of the technique is provided for two use cases: a cuffless blood pressure system in the form factor of a smartwatch (PPG signals acquired at the wrist working in reflection mode), and a cuffless blood pressure system in the form factor of a traditional fingertip sensor (PPG signals acquired at the fingertip working in transmission mode).
Clinical studies in collaboration with several university hospitals are currently ongoing to validate these techniques in multiple scenarios, from anesthetized patients in the operating room to hypertensive patients in ambulatory settings. Early evidence suggests that our pulse wave analysis approach can provide reliable continuous blood pressure measurements.

Full text


Abstract— While elevated blood pressure affects one third of
adults in developed countries, there is nowadays no comfortable
and easy-to-deploy technology to assess blood pressure in daily
life scenarios. The increasing use of smart devices has recently
spread the availability of photoplethysmographic (PPG) sensors
at everyone’s wrist (smartwatches), or eventually the finger
(camera-based sensing on smartphones). Although such PPG
sensors represent a strong opportunity to assess physiological
information on mobile device users, no reliable technique has so
far been demonstrated to exploit such PPG signals for
continuous cuffless blood pressure monitoring.
We present the results of a novel technique that provides
cuffless beat-to-beat estimates of blood pressure from PPG
signals with high level of accuracy. This optical technique relies
on the physiological analysis of PPG signals acquired either in
transmission or reflection mode. The developed algorithm
extracts features from PPG pulses based on the wave reflection
theory of the arterial system. The calculated features are further
converted into blood pressure values via a calibration procedure.
Performance results of the technique is provided for two use
cases: a cuffless blood pressure system in the form factor of a
smartwatch (PPG signals acquired at the wrist working in
reflection mode), and a cuffless blood pressure system in the
form factor of a traditional fingertip sensor (PPG signals
acquired at the fingertip working in transmission mode).
Clinical studies in collaboration with several university
hospitals are currently ongoing to validate these techniques in
multiple scenarios, from anesthetized patients in the operating
room to hypertensive patients in ambulatory settings. Early
evidence suggests that our pulse wave analysis approach can
provide reliable continuous blood pressure measurements.
I. A PULSE WAVE ANALYSIS ALGORITHM
A novel technique is presented to estimate BP of a subject
by analyzing PPG signals. After pre-processing the raw
signals, a pulse wave analysis (PWA) technique is applied to
each PPG pulse [1]. The implemented PWA algorithm
extracts information from each pulse based on the wave
reflection theory of the arterial system [2], and generates a
vector of physiologically meaningful features. Finally, a
calibration procedure is applied in order to translate
user-dependent changes of such PWA features into actual
changes of blood pressure, in mmHg.
J. Solà is with CSEM – Centre Suisse d’Electronique et de
Microtechnique, CH-2002, Neuchâtel, Switzerland, phone: +41 720 5112;
e-mail: Josep.Sola@csem.ch
M. Proença, F. Braun, E. Muntané-Calvo, C. Verjus, M. Lemay, M.
Bertschi, and J. Krauss are with CSEM – Centre Suisse d’Electronique et de
Microtechnique, CH-2002 Neuchâtel, Switzerland
P. Schoettker is with CHUV – Centre Hospitalier Universitaire Vaudois,
CH-1011 Lausanne, Switzerland, email: Patrick.Schoettker@chuv.ch
II. EXPERIMENTAL EVIDENCES
A. Reflection PPG: A Smartwatch Form Factor
Figure 1. provides experimental results of the performance of
the PWA algorithm when applied to PPG signals at the wrist.
Figure 1. Systolic BP estimation results from a smartwatch prototype vs
reference brachial cuff [3] based on a hand-grip protocol with 15 lying
subjects (different markers) aged 40±11 and BMI 22.9±2.7kg/m2.
B. Transmission PPG: A Fingertip Form Factor
Early experimental evidence of the performance of the novel
technique when applied to fingertip PPG signals for one
patient under general anesthesia is provided in Figure 2.
Figure 2. Systolic BP estimates from a fingertip sensor vs invasive systolic
measurements from an arterial catheter during general anesthesia [4].
III. CONCLUSIONS
Early evidence suggests that the analysis of PPG signals
via a dedicated PWA algorithm can provide the user with
reliable continuous blood pressure measurements.
REFERENCES
[1] Proença, Solà, Lemay & Verjus. PCT/EP 2015/063765. 2015.
[2] Nichols, O'Rourke & Vlachopoulos. ISBN 9780340985014
[3] Solà, Proença, Braun, Lemay & Verjus, CSEM Scientific. Rep. 2015
[4] Solà, Proença, Braun, et al, Proc. BMT2016
Cuffless Blood Pressure Monitoring:
Experimental Evidences of a Beat-to-Beat PPG Technique
J. Solà, M. Proença, F. Braun, E. Muntané-Calvo, C. Verjus,
M. Lemay, M. Bertschi, P. Schoettker, and J. Krauss