Available via license: CC BY
Content may be subject to copyright.
EDITED BY
Chenxi Yang,
Southeast University, China
REVIEWED BY
Yongxin Chou,
Changshu Institute of Technology, China
Chang Yan,
Southeast University, China
*CORRESPONDENCE
Christine Hove
christinehove2015@gmail.com
†
These authors have contributed equally to
this work and share first authorship
‡
These authors have contributed equally to
this work and share last authorship
RECEIVED 14 July 2024
ACCEPTED 01 October 2024
PUBLISHED 21 October 2024
CITATION
Hove C, Sæter FW, Stepanov A, Bøtker-
Rasmussen KG, Seeberg TM, Westgaard E,
Heimark S, Waldum-Grevbo B, Hisdal J and
Larstorp ACK (2024) A prototype
photoplethysmography-based cuffless device
shows promising results in tracking changes in
blood pressure.
Front. Med. Technol. 6:1464473.
doi: 10.3389/fmedt.2024.1464473
COPYRIGHT
© 2024 Hove, Sæter, Stepanov, Bøtker-
Rasmussen, Seeberg, Westgaard, Heimark,
Waldum-Grevbo, Hisdal and Larstorp. This is
an open-access article distributed under the
terms of the Creative Commons Attribution
License (CC BY). The use, distribution or
reproduction in other forums is permitted,
provided the original author(s) and the
copyright owner(s) are credited and that the
original publication in this journal is cited, in
accordance with accepted academic practice.
No use, distribution or reproduction is
permitted which does not comply with
these terms.
A prototype
photoplethysmography-based
cuffless device shows
promising results in tracking
changes in blood pressure
Christine Hove1,2*†, Frode Wirum Sæter1,3†, Alexey Stepanov4,
Kasper Gade Bøtker-Rasmussen4, Trine M. Seeberg4,
Espen Westgaard4, Sondre Heimark1,2, Bård Waldum-Grevbo1,2,
Jonny Hisdal1,3‡and Anne Cecilie K. Larstorp1,5,6‡
1
Institute of Clinical Medicine, University of Oslo, Oslo, Norway,
2
Department of Nephrology, Oslo
University Hospital, Oslo, Norway,
3
Department of Vascular Surgery, Oslo University Hospital, Oslo,
Norway,
4
Aidee Health AS, Bærum, Norway,
5
Section for Cardiovascular and Renal Research, Oslo
University Hospital, Oslo, Norway,
6
Department of Medical Biochemistry, Oslo University Hospital,
Oslo, Norway
Introduction: Non-invasive cuffless blood pressure devices have shown
promising results in accurately estimating blood pressure when comparing
measurements at rest. However, none of commercially available or prototype
cuffless devices have yet been validated according to the appropriate standards.
The aim of the present study was to bridge this gap and evaluate the ability of a
prototype cuffless device, developed by Aidee Health AS, to track changes in
blood pressure compared to a non-invasive, continuous blood pressure monitor
(Human NIBP or Nexfin) in a laboratory set up. The performance was evaluated
according to the metrics and statistical methodology described in the ISO
81060-3:2022 standard. However, the present study is not a validation study
and thus the study was not conducted according to the ISO 81060-3:2022
protocol, e.g., non-invasive reference and distribution of age not fulfilled.
Method: Data were sampled continuously, beat-to-beat, from both the cuffless
and the reference device. The cuffless device was calibrated once using the
reference BP measurement. Three different techniques (isometric exercise,
mental stress, and cold pressor test) were used to induce blood pressure
changes in 38 healthy adults.
Results: The mean difference (standard deviation) was 0.3 (8.7) mmHg for
systolic blood pressure, 0.04 (6.6) mmHg for diastolic blood pressure, and 0.8
(7.9) mmHg for mean arterial pressure, meeting the Accuracy requirement of
ISO 81060-3:2022 (≤6.0 (10.0) mmHg). The corresponding results for the
Stability criteria were 1.9 (9.2) mmHg, 2.9 (8.1) mmHg and 2.5 (9.5) mmHg. The
acceptance criteria for the Change requirement were achieved for the 85th
percentile of ≤50% error for diastolic blood pressure and mean arterial
pressure but were higher than the limit for systolic blood pressure (56% vs.
≤50%) and for all parameters for the 50th percentile (32%–39% vs. ≤25%).
Conclusions: The present study demonstrated that the cuffless device could track
blood pressure changes in healthy adults across different activities and showed
promising results in achieving the acceptance criteria from ISO 81060-3:2022.
KEYWORDS
cuffless, blood pressure, healthy adults, machine learning, blood pressure changes
TYPE Original Research
PUBLISHED 21 October 2024
|
DOI 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 01 frontiersin.org
1 Introduction
Cuffless, wearable blood pressure (BP) measurement devices
(cuffless devices) have the potential to provide continuous, beat-
to-beat BP estimations during daily routines, without significant
discomfort to the user (1,2). Despite considerable research in
this field, the accuracy of cuffless devices remains uncertain (3).
In these devices BP is estimated by device specific models, using
the input from physiological variables and signals that are related
to changes in BP. Most of them use pulse wave analysis of
photoplethysmographic (PPG) signals, pulse arrival time (PAT)
or a combination of both (4).
Several cuffless devices have been shown to accurately predict
BP in subjects at rest under controlled conditions in the
laboratory (5–8). Some cuffless devices are even commercialized
as validated according to the European Society of Hypertension
International Protocol Revision 2010 (5) and/or ISO 81060-
2:2019 protocol (6,9,10), which are not intended for cuffless
devices. There are several issues with validation of cuffless
devices using these protocols.
First, they are designed to test intermittent automated cuff-
based BP devices during static conditions over a short period of
time. In contrast to the cuff-based devices which aim to measure
the actual pressure, the cuffless BP devices provide surrogate BP
estimations from non-pressure signals and are prone to
fluctuations in these signals which are not related to BP (3,11,12).
Second, most cuffless BP devices rely on an initial, individual
calibration that is usually performed at rest using a standard
cuff-based BP device. Essentially these devices track changes in
BP relative to the calibration value (13). In a stable, resting
condition BP variations are small, and might be almost
negligible, especially when the duration of protocol is short. In
these cases, the device would seem to track BP accurately, but
this does not guarantee the same performance over longer
periods of time or under substantial BP changes (14).
Third, various situations commonly encountered in daily life,
such as physical activity, mental stress, and perception of
physical pain, produce changes in BP through different
physiological mechanisms. Thus, devices used in clinical
evaluation of BP must be able to accurately estimate BP changes
from a variety of activities (3).
To address these issues, several standards and recommendations
have been published recently. The ISO 81060-3:2022 standard (15)
focusesonvalidationofcuffless, noninvasive BP devices that
provide continuous, beat-to-beat or high-resolution BP estimations.
On the other hand, the ESH 2023 recommendations (3)describe
the validation procedure for intermittent cuffless devices for use
in ambulatory settings. To the best of our knowledge, no
cuffless device has yet been validated according to either the ISO
81060-3:2022 standard or to the ESH 2023 recommendations (3).
In contrast to most studies which evaluated the performance of
cuffless devices at rest, the aim of the present study was to address
these limitations and use three well-known techniques to alter BP
by different physiological mechanisms [isometric exercise (16),
mental stress (17,18), and cold pressor test (19–21)] to investigate
the ability of a prototype, PPG-based cuffless device, placed on the
upper arm, to track BP compared to a non-invasive, continuous
BP monitor in healthy adults. The present pilot study was
designed as part of the development of the cuffless device (Aidee
Health AS, Norway), towards a future validation. Thus, the
performance of the cuffless device was evaluated according to the
metrics from the ISO 81060-3:2022 standard (ISO 3).
2 Materials and methods
2.1 Participants
Healthy volunteers ≥18 years of age, free of any
chronic or cardiovascular disease, were eligible for inclusion.
Potential participants were screened with a short interview, BP
measurements (inclusion BP) and a 12-lead electrocardiogram
(ECG). Candidates with pregnancy, inclusion BP ≥180/120 mmHg
or any contraindication to standard cardiac stress testing (22)were
excluded. In line with the Helsinki declaration (23), all participants
were informed about the test procedure and signed a written
informed consent form before inclusion. The participants were
instructed to avoid intake of any food during the two hours prior to
the test, as well as caffeine drinks and nicotine during the four hours
before the test and alcohol at any time on the day of the test. During
the test the participants were dressed in comfortable clothes to
minimally interfere with the experimental conditions.
The study was approved by the Regional committees for
medical and health research ethics (REK, Norway, project
number 65844) prior to the inclusion of the first participant.
2.2 Reference blood pressure
Reference BP was measured continuously and non-invasively by
the volume-clamp method using either Human NIBP Nano System
(AD Instruments, Sydney, Australia) or Nexfin(24–30)(Bmeye,
Amsterdam, The Netherlands). Two different reference BP devices
were used as the Nexfin device malfunctioned during the study
and was replaced with the Human NIBP, which uses the same
measurement principle. The parent technology, Finapres (FMS
Finapres, Medical systems BV, Amsterdam) has been validated for
research use (31–33) and is commonly accepted for non-invasive
BP measurements in non-critically ill patients (17,34,35). The
finger pressure cuff was placed on the left middle finger. A laptop
was connected to the reference device, and the raw data was
sampled at 1,000 Hz, and continuously recorded during the
experiments using Lab Chart 8.1.9 software (AD Instruments,
Sydney, Australia). During each activity, the hand with the
reference device was maintained in a steady position to minimize
possible noise and artifacts. Between each activity there was a
pause where the reference device recording was stopped.
Therefore, the finger cuff device was calibrated at the start of each
activity by using a brachial cuff-based BP, which was measured on
the right upper arm with a validated automated oscillometric
device (Watch BP O3, Microlife Health Management Ltd.,
Cambridge, UK). Three readings separated by 1 min intervals were
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 02 frontiersin.org
taken during the 4-minute resting period at the beginning of each of
the three activities.
A 3-lead ECG was recorded continuously using Bio Amp/
PowerLab (AD Instruments) during the tests and the data were
exported to Lab Chart to calculate heart rate (HR).
Inclusion BP was measured on the right upper arm with the
participant in the supine position before the first activity using
the validated automated oscillometric device. Three consecutive
measurements were taken with 1 min intervals between
measurements. The first measurement was discarded, and the
average of the two remaining measurements were used to
calculate inclusion BP.
2.3 Cuffless blood pressure device
A prototype cuffless device, developed by Aidee Health AS
(Bærum, Norway), was used in the present study (Figure 1). The
device is the evolution of the technology that has been previously
described in several studies (2,36–39). It is a wearable device
with a PPG and an inertial measurement unit (consisting of 3D
accelerometer and gyroscope). Raw signals from the PPG sensor
were sampled at 1,000 Hz while accelerometer data was sampled
at 208 Hz and gyroscope data at 28 Hz. During the study the
device was placed on the left upper arm.
2.4 Study protocol
The study was conducted at the Department of Vascular
Surgery at Oslo University Hospital, Aker (Norway) from April
to October 2023. The protocol (Figure 2) consisted of three test
periods with three different activities to induce BP changes. Each
period consisted of a four-minute rest followed by the test
activity and 1-minute recovery. There was a longer rest of
5–10 min between each period.
Participants wore the cuffless device and the reference BP
device simultaneously. The first activity, isometric handgrip, was
performed by gripping the right hand around a custom-made
handgrip apparatus displaying the force applied by the
participant (16). Prior to the isometric handgrip, the maximal
voluntary contraction (MVC) force was measured. The
participants were instructed to keep 30% of MVC by looking at
the display during the two minutes of isometric handgrip, avoid
the Valsalva maneuver and relax all the muscles not primarily
involved in contraction. This was repeated three times with two-
minute pauses between each session of handgrip.
The second activity was a mental stress test where participants
subtracted 13 repetitively for five minutes starting with 1,079
(17,18). They were informed of any miscalculation in a direct
and stressful manner. A metronome at a frequency of two Hz
was used to distract and stress the participants.
The third activity was a cold pressor test (19–21) where the
right hand of the participant was completely immersed in ice
water (2–5°C) for two minutes.
Some of the participants wore the cuffless device for 24 h after
the laboratory tests in order to test the stability of the cuffless BP
estimations. The participants did not wear the reference device
outside of the laboratory. On the following day, we repeated the
isometric handgrip test with the participant wearing both the
cuffless device and the reference device.
FIGURE 1
Illustration of the cuffless blood pressure device.
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 03 frontiersin.org
2.5 Data processing
Filtering and processing of the data was performed post-hoc by
using Python programming language.
2.5.1 Reference blood pressure
Reference BP values were calculated from the recorded raw BP
waveforms and calibrated using brachial systolic BP (SBP)
measurement. The brachial measurement was the mean of the
last two of three measurements measured in the rest period
before each activity. The raw BP waveforms were then shifted to
align the peaks with the calibration measurement.
The raw waveform signals were automatically filtered to
remove artefacts, such as periods of automatic calibration
(AutoCal/Physiocal) and high frequency noise. Then for each
cardiac cycle, defined by R-peaks in the ECG signal, the systolic,
diastolic (DBP), and mean arterial pressure (MAP) were
computed using maximum, minimum and time-weighted integral
correspondingly. Then, all data was controlled manually for
artefacts by comparing systolic and diastolic values with peaks
and by reviewing actual BP waves for every subject. Finally,
mean SBP, DBP and MAP were calculated per non-overlapping
15-second segments.
Participants were excluded from the statistical analyses if
more than 50% of their reference data had to be removed due to
artefacts or noise.
2.5.2 Cuffless device
The raw PPG signals from the cuffless device were processed
and filtered using proprietary algorithms. The signals were
divided into cardiac cycles and averaged over non-overlapping
15-second segments. Segments with unacceptable data quality,
i.e., artefacts or noise, were removed. For each 15-second
segments multiple standard features from the PPG signal
commonly presented in the literature (40–42) were extracted.
2.5.3 Cuffless blood pressure models and
calibration
The cuffless BP models were developed from the present study
cohort using 3-fold Cross-validation (43–45), a statistical method
to evaluate the performance of the model in case of limited
data (46). Separate models were made for each BP parameter
(i.e., SBP, DBP and MAP) using the following procedure: First,
the subjects were split into three subsets (“folds”), which were
then used to train three different regression models for each BP
parameter (Figure 3). Then a final model for each BP parameter
was derived (based on averaging the three regression models).
The final three models were then used to predict SBP, DBP and
MAP separately.
Contrary to the reference BP, the cuffless device was only
calibrated once to correct the offset between reference BP and
cuffless BP. This was done during the initial rest period before the
handgrip activity using the calibrated reference BP value.
Participants’demographics were not used for additional calibration.
2.6 Statistical analyses
Statistical analysis was performed using Stata 18.0 (Statacorp.,
Texas, USA). Variables were assessed for normality by visual
inspection of histograms. Continuous data are presented as mean
(standard deviation; SD), or median (interquartile range; IQR) if
non-normally distributed. For each participant, within-subject
FIGURE 2
Illustration of the test protocol with activities and rest periods. Created in BioRender. Sæter, F. (2024) BioRender.com/r16m233.
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 04 frontiersin.org
change in BP and HR was calculated by taking the highest reported
reference BP or HR measurement subtracted by the lowest reported
reference BP measurement during the entire test period day 1
(handgrip, mental stress test, cold pressor test) and for each
activity separately.
We chose to adopt, as closely as possible given the differences
in protocol, the same statistical methodology as described in ISO
3. The ISO 3 includes three requirements for evaluating
performance of cuffless devices: (1) the Accuracy criteria
[Chapter 5.1 (15)], (2) the Stability criteria [Chapter 5.2 (15)]
and (3) the Change criteria [Chapter 5.3 (15)]. We have
compared our results against the acceptance criteria for all three
tests. The acceptance criteria for the Accuracy and Stability
requirements from ISO 3 is a mean difference (SD) ≤6
(10) mmHg. The acceptance criterion for the Change
requirement is two-folded: averaged calculated (1) 50th percentile
of error rate between the reference device and the cuffless device
for the specified change evaluation interval ≤25%, and (2) 85th
percentile ≤50%. To evaluate the Accuracy and Change criteria,
we used the data collected from the whole test period of the first
day (handgrip, mental stress test and cold pressor test). For the
BP change parameters included in the Change analysis, the start/
end points for BP change were kept within the same activity
(either handgrip, mental stress test or cold pressor test). In the
data analysis for the Stability criteria, we used data from the
participants included in the 24 h test: the data collected during
the first day’s test period (handgrip, mental stress test and cold
pressor test) and during the test period of the following day
(handgrip day 2).
In addition, we evaluated the level of absolute agreement
between the reference BP device and the cuffless device for SBP,
DBP and MAP, during the entire first day, using Bland-Altman
plots with bias and 95% limits of agreement (LoA). We
acknowledge that aggregating all measurement pairs across all
patients may violate the assumption of independent measurements
in the Bland-Altman method (47). However, most cuffless studies
have adopted this approach in their analyses (48–52).
3 Results
3.1 Participant selection, general
characteristics and blood pressure
distribution
A total of 67 participants were recruited, of whom 29 were
excluded due to unacceptable noise in the reference BP data.
Thus, 38 participants were included in the statistical analyses.
General characteristics for the cohort are presented in Table 1.
The BP range for each individual during the entire test protocol
is presented in Figure 4. Reference BP and HR distribution
during the test protocol are presented in Table 2.
3.2 Performance of the cuffless blood
pressure model
To determine the minimum number of repeated paired
measurements and number of subjects, the intraclass correlation
coefficient (ICC) was estimated a priori as outlined in ISO
3. Post-hoc the ICC, that was calculated from the reference data
included in the analysis, was 0.2 for SBP, 0.3 for DBP and 0.2
for MAP for the Accuracy analysis, and 0.3 for SBP, 0.3 for DBP
and 0.3 for MAP for the Stability analysis.
Twenty-two randomly chosen pairwise comparisons between
reference and cuffless BP per subject, i.e., a total of 836
measurement pairs for each BP parameter, were used to evaluate
the Accuracy criteria. Forty-four randomly chosen pairwise
comparisons between reference and cuffless BP per subject, i.e., a
total of 484 measurement pairs for each BP parameter, were
included in the Stability analysis. A total of 3,549 measurement
pairs for SBP, 3,142 for DBP and 3,510 for MAP were included
in the Change analysis.
Table 3 summarizes the comparison between the cuffless device
and the reference BP device, with respect to the acceptance criteria
FIGURE 3
Illustration of the 3-fold cross validation procedure.
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 05 frontiersin.org
outlined in ISO 3. The mean difference (SD) was 0.3 (8.7) mmHg
for SBP, 0.04 (6.6) mmHg for DBP, and 0.8 (7.9) mmHg for MAP
for the Accuracy requirement. The corresponding mean differences
for the Stability requirements were 1.9 (9.2), 2.9 (8.1), and 2.5
(9.5) mmHg for SBP, DBP and MAP, respectively. Thus, all BP
parameters were within acceptance criteria for the Accuracy and
Stability requirements (≤6.0 (10.0) mmHg).
The 50th and 85th percentile of error rate between the
reference device and the cuffless device was 39% and 56% for
SBP, 32% and 48% for DBP and 33% and 47% for MAP,
respectively. Thus, the cuffless device achieved the acceptance
criteria for the Change requirement for the 85th percentile of
≤50% error for DBP and MAP but were higher than acceptable
for SBP (55% vs. ≤50%) and for all parameters for the 50th
percentile of error rate (32%–39% vs. ≤25%).
To exemplify the results, we included time series plots from
six selected participants in Figure 5.Notethatbecausethe
reference device was calibrated for each activity, while the
cuffless device was calibrated only once, there is a notable
offset between BP readings for certain participants and
activities (see Figure 5). This does not influence results for
Change, but the Accuracy and Stability metrics could
potentially be improved if we had not recalibrated the
reference BP before each activity.
The degree of agreement between the cuffless device and
reference device during the entire test protocol day 1 is presented
with Bland Altman plots with bias and 95% LoA (Figure 6). Bias
[95% LoA] was close to zero for all BP parameters over the
entire test period day 1 (all activities); 0.24 mmHg [−8.7,
9.2 mmHg], 0.63 mmHg [−7.3, 8.5 mmHg] and 0.77 mmHg
[−7.2, 8.7 mmHg] for SBP, DBP and MAP, respectively.
4 Discussion
The present study aimed to evaluate the ability of a prototype,
PPG-based cuffless device, placed on the upper arm, to track BP
during three well-known activities to induce BP changes using
different physiological mechanisms. The results demonstrated
that the cuffless device estimated BP with satisfactory accuracy
compared to a non-invasive, continuous reference BP monitor, in
38 healthy adults in an experimental laboratory set up consisting
of isometric handgrip (16), mental stress test (17,18) and cold
pressor test (19–21). The cuffless device showed promising
results in achieving the acceptance criteria from the ISO 81060-
3:2022 standard (ISO 3) (15). To the best of our knowledge, this
is the first study to present results according to the full statistical
methodology outlined in the ISO 3, which is the first standard
addressing the validation of cuffless devices.
The cuffless device fulfilled the acceptance criteria (≤6
(10) mmHg) for the Accuracy and Stability requirements from
ISO 3. A particular strength of our study is that these metrics
were calculated using the whole measurement period, including
the periods with the induced BP change (unstable periods) which
is not required by ISO 3. Even though the present study is not a
validation study, we adopted as closely as possible (given our
different protocol), the metrics and statistical methodology from
ISO 3, which addresses mentioned issues with evaluating
performance of cuffless devices and represents state-of-the-art
benchmark for the present and similar studies.
A few other cuffless devices have demonstrated the ability to
accurately estimate BP in individuals at rest during stable
conditions. A study evaluating a PPG-based cuffless device,
worn as a bracelet (Aktiia), compared to auscultation in
FIGURE 4
Box plot of blood pressure distribution for each individual.
TABLE 1 General characteristics of the included participants (n= 38).
Number (%) 38 (100)
Handgrip (day 1) 38 (100)
Mental stress test (day 1) 36 (94.7)
Cold pressor test (day 1) 33 (86.8)
24 h test (handgrip day 2) 11 (28.9)
Age, median (IQR), years
Number (%)
33 (20)
•>50 years
•>60 years
•>70 years
5 (13.2)
2 (5.3)
0 (0)
Female sex, number (%) 22 (57.9)
Body mass index, mean (SD), kg/m
2
23.4 (2.8)
Baseline Systolic Blood Pressure (supine position), mean (SD),
mmHg
119.3 (9.4)
Baseline Diastolic Blood Pressure (supine position), mean (SD),
mmHg
72.6 (7.1)
Fitzpatrick skin pigmentation, no (%)
1 5 (13.2)
2 27 (71.1)
3 4 (10.5)
4 1 (2.6)
5 1 (2.6)
6 0 (0)
7 0 (0)
IQR, interquartile range; SD, standard deviation.
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 06 frontiersin.org
91 adults (6) demonstrated accurate BP predictions in the seated,
supine and standing position with a mean difference (SD) for SBP
of 0.5 (7.8) mmHg in the sitting position, −2.4 (10.1) mmHg in
the supine, and −0.6 (12.5) mmHg in the standing position.
Differences for DBP readings were 0.4 (6.9) mmHg, −1.9
(7.7) mmHg, and −4.9 (9.1) mmHg respectively. Accuracy of the
same device (Aktiia) was compared to auscultation in 35 elderly
individuals in the seated, supine and standing position (8)and
demonstrated similar results. Another study evaluating a cuffless
device that is based on pulse transit time (Somnotouch-NIBP)
demonstrated similar results in 33 subjects in the seated
position (5). The BP estimations of a cuffless, wrist-worn or
skin attached device (BioBeat), that uses pulse wave analysis
of the PPG signal in combination with pulse wave transit
times, was compared to the measurements of a standard
sphygmomanometer device in 1,057 subjects in the seated
position (7). In this study the BioBeat device was found similar
to the sphygmomanometer device with high agreement and
reliability levels. However, none of these studies have shown that
cuffless devices can accurately track substantial physiological BP
changes. This is an important aspect, as cuffless devices only
track changes in BP relative to the calibration value. Thus, in a
stable, resting condition where BP variations are small, a device
would seem to track BP accurately even though this might not
be true under larger BP changes.
Unlike most studies on cuffless devices, which predominantly
focus on accuracy assessment during resting conditions, we
evaluated performance of the cuffless device during large BP
changes induced by three different physiological mechanisms, i.e.,
isometric handgrip, mental stress test, and cold pressor test. The
effects of isometric exercise on the cardiovascular system were first
described by Lindhard in 1920 (54). Since then, it has been shown
that isometric exercise causes a concurrent increase in both SBP
and DBP (36,55–57). The BP response to mental stress is
characterized by a predominant elevation in SBP, reflecting
increased cardiac output driven by an increase in both stroke
volume and HR, while DBP may remain relatively stable or show
amodestincrease(58,59). Cold induced pain typically results in a
rapid and consistent BP elevation during the stimulus due to an
immediate sympathetic surge, primarily in SBP, while DBP may
also rise (21,60,61). Despite using these different mechanisms to
induce BP changes, we still demonstrated high agreement in the
ability of the cuffless BP device to track SBP, DBP and MAP.
Furthermore, the cuffless device showed promising results in
meeting the acceptance criteria for the Change requirement of
ISO 3 (50th percentile ≤25% and 85th percentile ≤50%). The
85th percentile of error rate for MAP and DBP was within the
acceptance criteria but was higher than the limit for SBP for the
85th percentile and for all BP parameters for the 50th percentile.
Only one comparative study has presented results partially
TABLE 2 Blood pressure and heart rate distribution of all individual measurements during the entire test protocol.
Systolic blood
pressure, mmHg
Diastolic blood
pressure, mmHg
Mean arterial
pressure, mmHg
Heart rate, beats
per minute
All activities day 1
Range, min–max 96–221 38–122 57–153 41–108
Within-subject change, median (IQR) 41.3 (26.5) 27.1 (10.7) 32.1 (13.1) 27.5 (10.1)
Handgrip (day 1)
Range, min–max 96–221 40–122 59–153 42–96
Within-subject change, median (IQR) 25.5 (22.6) 17.1 (11.6) 21.7 (13.5) 15.8 (8.5)
Mental stress test (day 1)
Range, min–max 106–198 38–108 57–135 45–108
Within-subject change, median (IQR) 22.6 (15.1) 13.0 (9.0) 16.2 (11.7) 21.6 (13.6)
Cold pressor test (day 1)
Range, min–max 98–201 42–119 62–149 41–100
Within-subject change, median (IQR) 35.1 (19.0) 21.0 (12.2) 27.3 (15.0) 14.1 (8.1)
Handgrip day 2
Range, min–max 105–196 39–104 60–138 44–98
Within-subject change, median (IQR) 33.3 (23.1) 19.1 (17.2) 23.6 (19.1) 15.7 (7.7)
IQR, interquartile range.
TABLE 3 Performance of the cuffless blood pressure device in comparison to the ISO 81060-3:2022 acceptance criteria.
Accuracy
Mean Δ(SD),
mmHg
Stability
a
Mean Δ(SD),
mmHg
ISO Criteria for
Accuracy and Stability
Mean Δ(SD), mmHg
Change
50th and 85th
percentile, %
ISO Criteria for
Change 50th and
85th percentile, %
Systolic blood pressure 0.3 (8.7) 1.9 (9.2)
a
≤6.0 (10.0) 39, 56 ≤25,≤50
Diastolic blood pressure 0.04 (6.6) 2.9 (8.1)
a
≤6.0 (10.0) 32, 48 ≤25,≤50
Mean arterial pressure 0.8 (7.9) 2.5 (9.5)
a
≤6.0 (10.0) 33, 47 ≤25,≤50
a
Only 11 participants were included in the Stability analysis.
Δ, difference; SD, standard deviation.
Text and values in Italic font indicate the pass criteria of the ISO 81060-3:2022 standard.
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 07 frontiersin.org
according to the statistical methodology outlined in ISO 3, i.e.,
Khayat et al. recently evaluated a wearable sensor against intra-
arterial BP measurements for the Change criteria in 27 patients
undergoing surgery, achieving a 50th percentile and 85th
percentile of error rate of 23.8% and 42%, respectively (62),
meeting the ISO 3 Change criteria (≤25% and ≤50% error for
the 50th and 85th percentiles respectively). Even though the
cuffless device, tested in the present study, only partially fulfilled
the acceptance criteria for the Change requirement, we argue that
the results are promising towards meeting the criteria in a future
validation study for several reasons. First, ISO 3 only requires a
limited increase (15 mmHg in SBP, 10 mmHg in DBP and
12 mmHg in MAP) for the BP change included in the Change
analysis, and it does not require comparison of measurements
obtained during this period of change where BP is unstable. In
the present study, we induced substantial BP changes in our
FIGURE 5
Time series plots of the results from the entire test protocol day 1 from six different participants for reference and cuffless systolic blood pressure
(SBP), diastolic blood pressure (DBP) and mean arterial pressure (MAP). The y-axis represents blood pressure (mmHg) and x-axis time (minutes).
The results are from three subjects with good agreement (A–C) and three subjects with mediocre agreement (D–F).
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 08 frontiersin.org
participants, and all available pairs were included in the Change
analysis. During some activity periods BP was changing
extremely fast. This might have introduced a higher uncertainty
in BP measurements from both the reference and cuffless device.
Thus, our results may have been better had we excluded periods
with unstable BP. However, we decided not to do this to clearly
illustrate the ability of the technology to track changes in
challenging conditions as well. Second, ISO 3 only requires a
certain BP change but does not specify how this BP change shall
be achieved. Alterations in BP can be induced by different
mechanisms and stimuli, and the hemodynamic responses to
these can vary in terms of magnitude and duration. In the
present study, we used 3 different interventions to induce BP
changes through different physiological mechanisms, instead of
just using one single exercise that is typically performed (36,37,
63). Third, the activities used to induce BP changes were equally
weighted in our Change calculations. The cold pressor test,
where the right hand of the participant was completely immersed
in ice water, could in some individuals have led to a significant
peripheral vasoconstriction in the extremity contralateral to the
cold immersion (64) and probably introduced a higher
uncertainty in the BP measurements conducted by the reference
device (64,65). We believe that these factors may explain why
we did not fulfill all acceptance criteria outlined in ISO 3 in the
present study and argue that the cuffless device showed
promising results in meeting the Change requirements in a
future validation study.
5 Deviances from the ISO
81060-3:2022 protocol
Even though we used methodology from ISO 81060-3:2022 to
evaluate our results, it is important to note that the protocol used in
the present study differs from the protocol described in ISO 3.
Most importantly, we used a different reference method. The
ISO 3 protocol requires an intra-arterial BP reference. This
involves cannulation of a peripheral artery, most commonly the
radial artery, with a catheter. In the present study, the Human
NIBP and Nexfin, which deliver continuous BP readings via a
non-invasive dual finger cuff system, were used as reference BP
devices. Both instruments use the volume-clamp methodology to
assess arterial pressure in the finger and by that calculate BP
(66). The accuracy of the Human NIBP Nano is according to the
manufacturer ± 1% of the full range (max. 3 mmHg) (65). Nexfin
has been compared to intra-arterial measurements in several
studies, demonstrating bias (SD) ranging from −1.2 (6.5) mmHg
to −4.6 (6.5) mmHg (27–30). The parent technology, Finapres
(FMS Finapres, Medical systems BV, Amsterdam), has been
demonstrated to be accurate when compared to intra-arterial
pressure with only minor discrepancies (17,34) and has been
validated for use in research (31,32). Finapres has proven to be
reliable in monitoring BP during dynamic changes (67,68).
However, a few studies have shown these and comparable devices
to be less accurate than intra-arterial BP measurements (69),
especially for SBP (33,70), and they are not recommended for
hemodynamic monitoring in critically ill patients where sudden
hypotension may occur (71). Nevertheless, the finger cuff devices
are commonly accepted as reliable for non-invasive BP
measurements in non-critically ill patients (17,34,35). While the
use of intra-arterial measurements provides enhanced accuracy, it
requires an invasive procedure, thereby also raising ethical,
practical, and financial considerations. Therefore, for our healthy
population the volume-clamp device was considered adequate
and appropriate.
Second, the total test period for each subject did not fulfill the
requirement for the Stability requirement of ISO 3. The standard
requires measurements during the first 5 h and again after 24 h
for devices intended for 24 h monitoring. This was not feasible
for the present study.
Third, the number of test subjects was lower than required for
the Stability requirement. For the given ICC in our dataset, ISO 3
requires at least 30 test subjects. We included more than an
adequate number of test subjects (n= 38) for the Accuracy and
Change analysis. However, only 11 individuals completed the
24 h measurement providing data for the Stability analysis.
FIGURE 6
Bland-Altman plots for individual blood pressure (BP) readings for each BP parameter day 1 of the test protocol. Mean of BP values from the reference
device and cuffless device (x-axis) plotted against the difference between reference and cuffless BP values (y-axis). Horizontal red lines indicate bias
and horizontal blue lines indicate upper and lower 95% limits of agreement. Vertical, dotted lines represent mean (±2 SD). Outliers, defined as BP
differences above 30 mmHg and below −30 mmHg (53), are plotted at point 30 mmHg and −30 mmHg, respectively.
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 09 frontiersin.org
Fourth, the subject characteristics, including the distribution of
BP, did not fulfil the criteria of ISO 3. Our cohort was relatively
young, i.e., only 13% >50 years (requirement >40%), 5% >60 years
(requirement >25%) and none >70 years (requirement >10%).
6 Strengths and limitations
A strength of the present study is that we evaluated performance
of the cuffless device during relatively large BP changes. This aspect
is essential, given that cuffless devices estimate BP changes relative to
a calibration value. Consequently, under stable, resting settings where
BP fluctuations are minimal, a device may appear to track BP
accurately. However, this perceived accuracy may not hold true
when there are larger variations in BP. Furthermore, we used three
well-known activities to induce BP changes via different
physiological mechanisms, i.e., isometric handgrip, mental stress
test, and cold pressor test.
Another strength is that we assessed performance of the
cuffless device in accordance with the statistical requirements
from the ISO 81060-3:2022 standard. This standard provides a
framework and statistics that enables direct comparisons between
different continuous, cuffless devices during both stable and
unstable conditions. Even though the present study is not a
validation study, using these metrics is particularly meaningful
for future comparisons with other devices. Additionally, we have
highlighted our results along with important deviances from the
ISO 81060-3:2022 protocol, which we believe are noteworthy for
future investigations and device comparisons.
However, the present study has several limitations. First, the study
was conducted under highly controlled laboratory conditions.
Consequently, the results may not be directly applicable to real-life
settings. Second, most of our participants had light skin color
(Fitzpatrick 1–3). Third, we only included healthy individuals.
Cuffless BP devices make presumptions about the arterial pulse
wave in their BP prediction models. Thus, the BP prediction
models might not be generalized to individuals with chronic
diseases, such as peripheral artery disease or cardiovascular diseases,
pregnant women, individuals with obesity, darker skin tones and/or
tattooed skin (14). The aim of the present study was to induce
relatively large BP changes. Thus, for ethical reasons we chose to
exclude candidates with comorbidities, such as hypertension and
cardiovascular diseases. Further research is necessary to determine
device performance in these sub-populations, which are currently
under-represented in clinical trials.
7 Conclusions
The present study demonstrated the ability of a prototype,
photoplethysmography-based cuffless device, placed on the upper
arm, to track large BP changes induced by different physiological
mechanisms. The cuffless device showed promising results in
achieving the acceptance criteria for the Accuracy, Stability and
Change requirements from the ISO 81060-3:2022 standard in
healthy adults. However, it is important to note, that we used a
different reference BP device and did not fulfill the requirement
for participant characteristics. Furthermore, the subject number
and study protocol time was not in accordance with the standard
for the Stability requirement.
The results of the present study are optimistic towards the
clinical use of cuffless devices in BP monitoring in healthy
adults. However, further research and validation is needed before
the technology can be implemented in health care.
Data availability statement
BP predictions from the cuffless BP model and the reference
measurements can be made available upon a reasonable request.
Raw signals and data regarding model development may not
be disclosed.
Ethics statement
The studies involving humans were approved by the Regional
committees for medical and health research ethics (REK,
Norway, project number 65844). The studies were conducted in
accordance with the local legislation and institutional
requirements. The participants provided their written informed
consent to participate in this study.
Author contributions
CH: Conceptualization, Data curation, Formal Analysis,
Investigation, Methodology, Writing –original draft, Writing –
review & editing, Visualization. FS: Conceptualization, Data
curation, Formal Analysis, Investigation, Methodology, Writing –
original draft, Writing –review&editing,Visualization.AS:
Conceptualization, Data curation, Formal Analysis, Methodology,
Writing –review & editing. KB-R: Conceptualization, Data
curation, Formal Analysis, Methodology, Writing –review &
editing. TS: Conceptualization, Data curation, Formal Analysis,
Funding acquisition, Methodology, Project administration,
Supervision, Writing –review & editing. EW: Conceptualization,
Funding acquisition, Methodology, Project administration, Software,
Writing –review & editing. SH: Conceptualization, Methodology,
Supervision, Writing –review & editing. BW-G: Conceptualization,
Methodology, Project administration, Resources, Supervision,
Writing –review & editing. JH: Conceptualization, Methodology,
Project administration, Resources, Supervision, Writing –review &
editing. AL: Conceptualization, Methodology, Project administration,
Resources, Supervision, Writing –review & editing.
Funding
The author(s) declare financial support was received for the
research, authorship, and/or publication of this article. The research
project (Hypersension 2.0, 2022-2026) is funded by the BIA
program of the Norwegian research council (project number 332371).
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 10 frontiersin.org
Acknowledgments
The authors thank statistician and researcher Lien My Diep at
Oslo Centre for Biostatistics & Epidemiology, Research Support
Services, Oslo University Hospital, Oslo, Norway, for guidance
with the statistical analysis.
Conflict of interest
AS, ES, KB-R and TS are employees of the company behind the
prototype cuffless multi-sensor device.
The remaining authors declares that the research was
conducted in the absence of any commercial or financial
relationships that could be construed as a potential conflict
of interest.
Publisher’s note
All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed
or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found
online at: https://www.frontiersin.org/articles/10.3389/fmedt.2024.
1464473/full#supplementary-material
SUPPLEMENTARY FIGURE S7
Histogram of all reference blood pressure readings (n= 836) included in the
ISO Accuracy analysis for each blood pressure parameter.
SUPPLEMENTARY FIGURE S8
Scatter plot of all reference blood pressure readings (n= 836) included in the
ISO Accuracy analysis for each blood pressure parameter. The y-axis
represents included reference blood pressure (mmHg), x-axis time
(minutes) from calibration of the cuffless device.
SUPPLEMENTARY FIGURE S9
Histogram of all reference blood pressure readings (n= 484) included in the
ISO Stability analysis for each blood pressure parameter.
SUPPLEMENTARY FIGURE S10
Histogram of all reference blood pressure readings (n= 7,098 for systolic, n
= 6,284 for diastolic and n= 7,020 for mean arterial pressure) include d in the
ISO Change analysis for each blood pressure parameter.
SUPPLEMENTARY FIGURE S11
Histogram of all reference blood pressure change parameters (n= 3,549 for
systolic, n= 3,142 for diastolic and n= 3,510 for mean arterial pressure)
included in the ISO Change analysis for each blood pressure parameter.
The reference blood pressure change parameter was calculated by taking
the reported reference blood pressure measurement at the end of the
blood pressure change minus the reported reference blood pressure
measurement at the start of the blood pressure change.
References
1. Bradley CK, Shimbo D, Colburn DA, Pugliese DN, Padwal R, Sia SK, et al.
Cuffless blood pressure devices. Am J Hypertens. (2022) 35(5):380–7. doi: 10.1093/
ajh/hpac017
2. Heimark S, Hove C, Stepanov A, Boysen ES, Gløersen Ø, Bøtke-Rasmussen KG,
et al. Accuracy and user acceptability of 24-hour ambulatory blood pressure
monitoring by a prototype cuffless multi-sensor device compared to a conventional
oscillometric device. Blood Press. (2023) 32(1):2274595. doi: 10.1080/08037051.2023.
2274595
3. Stergiou GS, Avolio AP, Palatini P, Kyriakoulis KG, Schutte AE, Mieke S, et al.
European Society of hypertension recommendations for the validation of cuffless
blood pressure measuring devices: European society of hypertension working group
on blood pressure monitoring and cardiovascular variability. J Hypertens. (2023) 41.
doi: 10.1097/HJH.0000000000003483
4. Mukkamala R, Shroff SG, Landry C, Kyriakoulis KG, Avolio AP, Stergiou GS. The
microsoft research aurora project: important findings on cuffless blood pressure
measurement. Hypertension. (2023) 80(3):534–40. doi: 10.1161/
HYPERTENSIONAHA.122.20410
5. Bilo G, Zorzi C, Ochoa Munera JE, Torlasco C, Giuli V, Parati G. Validation of
the somnotouch-NIBP noninvasive continuous blood pressure monitor according to
the European society of hypertension international protocol revision 2010. Blood
Press Monit. (2015) 20(5):291–4. doi: 10.1097/MBP.0000000000000124
6. Sola J, Vybornova A, Fallet S, Polychronopoulou E, Wurzner-Ghajarzadeh A,
Wuerzner G. Validation of the optical aktiia bracelet in different body positions for
the persistent monitoring of blood pressure. Sci Rep. (2021) 11(1):20644. doi: 10.
1038/s41598-021-99294-w
7. Nachman D, Gepner Y, Goldstein N, Kabakov E, Ishay AB, Littman R, et al.
Comparing blood pressure measurements between a photoplethysmography-based
and a standard cuff-based manometry device. Sci Rep. (2020) 10(1):16116. doi: 10.
1038/s41598-020-73172-3
8. Theiler K, Sola J, Damianaki A, Pfister A, Almeida TP, Alexandre J, et al.
Performance of the aktiia optical blood pressure measurement device in the elderly:
a comparison with double blinded auscultation in different body positions. Blood
Press. (2023) 32(1):2281320. doi: 10.1080/08037051.2023.2281320
9. Vybornova A, Polychronopoulou E, Wurzner-Ghajarzadeh A, Fallet S, Sola J,
Wuerzner G. Blood pressure from the optical aktiia bracelet: a 1-month validation
study using an extended ISO81060-2 protocol adapted for a cuffless wrist device.
Blood Press Monit. (2021) 26(4):305–11. doi: 10.1097/MBP.0000000000000531
10. Schoettker P, Degott J, Hofmann G, Proença M, Bonnier G, Lemkaddem A, et al.
Blood pressure measurements with the OptiBP smartphone app validated against
reference auscultatory measurements. Sci Rep. (2020) 10(1):17827. doi: 10.1038/
s41598-020-74955-4
11. Stergiou GS, Mukkamala R, Avolio A, Kyriakoulis KG, Mieke S, Murray A, et al.
Cuffless blood pressure measuring devices: review and statement by the European
society of hypertension working group on blood pressure monitoring and
cardiovascular variability. J Hypertens. (2022) 40(8):1449–60. doi: 10.1097/HJH.
0000000000003224
12. Khan Mamun MMR, Sherif A. Advancement in the cuffless and noninvasive
measurement of blood pressure: a review of the literature and open challenges.
Bioengineering (Basel). (2022) 10(1):27. doi: 10.3390/bioengineering10010027
13. Mukkamala R, Yavarimanesh M, Natarajan K, Hahn JO, Kyriakoulis KG, Avolio
AP, et al. Evaluation of the accuracy of cuffless blood pressure measurement devices:
challenges and proposals. Hypertension. (2021) 78(5):1161–7. doi: 10.1161/
HYPERTENSIONAHA.121.17747
14. Hu J-R, Martin G, Iyengar S, Kovell LC, Plante TB, Nv H, et al. Validating
cuffless continuous blood pressure monitoring devices. Cardiovasc Digit Health J.
(2023) 4(1):9–20. doi: 10.1016/j.cvdhj.2023.01.001
15. Technical Committee ISO/TC 121, Anaesthetic and respiratory equipment,
Subcommittee SC 3, Respiratory devices and related equipment used for patient
care, and Technical Committee IEC/TC 62, Electrical equipment in medical
practice, Subcommittee SC 62D, Electromedical equipment, in collaboration with
the European Committee for Standardization (CEN) Technical Committee CEN/TC
205, Non-active medical devices. ISO 81060-3:2022 Non-invasive
Sphygmomanometers Part 3: Clinical Investigation of Continuous Non-invasive
Automated Meaureument Type. Released 12/16/2022.
16. Bakke EF, Hisdal J, Kroese AJ, Jørgensen JJ, Stranden E. Blood pressure response
to isometric exercise in patients with peripheral atherosclerotic disease. Clin Physiol
Funct Imaging. (2007) 27(2):109–15. doi: 10.1111/j.1475-097X.2007.00720.x
17. Rostrup M, Westheim A, Kjeldsen SE, Eide I. Cardiovascular reactivity, coronary
risk factors, and sympathetic activity in young men. Hypertension. (1993) 22(6):891–9.
doi: 10.1161/01.HYP.22.6.891
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 11 frontiersin.org
18. Schmieder R, Rüddel H, Schãchinger H, Neus H. How to perform mental stress
tests. J Hum Hypertens. (1987) 1(3):223–8.
19. Rostrup M, Kjeldsen SE, Eide IK. Awareness of hypertension increases blood
pressure and sympathetic responses to cold pressor test. Am J Hypertens. (1990) 3
(12 Pt 1):912–7. doi: 10.1093/ajh/3.12.912
20. Hines EA, Brown GE. The cold pressor test for measuring the reactibility of the
blood pressure: data concerning 571 normal and hypertensive subjects. Am Heart J.
(1936) 11(1):1–9. doi: 10.1016/S0002-8703(36)90370-8
21. Mourot L, Bouhaddi M, Regnard J. Effects of the cold pressor test on cardiac
autonomic control in normal subjects. Physiol Res. (2009) 58(1):83–91. doi: 10.
33549/physiolres.931360
22. Fletcher GF, Ades PA, Kligfield P, Arena R, Balady GJ, Bittner VA, et al. Exercise
standards for testing and training: a scientific statement from the American Heart
Association. Circulation. (2013) 128(8):873–934. doi: 10.1161/CIR.0b013e31829b5b44
23. World Medical Association. World medical association declaration of Helsinki:
ethical principles for medical research involving human subjects. JAMA. (2013) 310
(20):2191–4. doi: 10.1001/jama.2013.281053
24. Ameloot K, Van De Vijver K, Van Regenmortel N, De Laet I, Schoonheydt K,
Dits H, et al. Validation study of nexfin
®
continuous non-invasive blood pressure
monitoring in critically ill adult patients. Minerva Anestesiol. (2014) 80(12):1294–301.
25. Eeftinck Schattenkerk DW, van Lieshout JJ, van den Meiracker AH, Wesseling
KR, Blanc S, Wieling W, et al. Nexfin noninvasive continuous blood pressure validated
against riva-rocci/korotkoff. Am J Hypertens. (2009) 22(4):378–83. doi: 10.1038/ajh.
2008.368
26. Pouwels S, Lascaris B, Nienhuijs SW, Arthur Bouwman R, Buise MP. Validation
of the nexfin
®
non-invasive continuous blood pressure monitoring validated against
riva-rocci/korotkoff in a bariatric patient population. J Clin Anesth. (2017)
39:89–95. doi: 10.1016/j.jclinane.2017.03.029
27. Martina JR, Westerhof BE, van Goudoever J, de Beaumont EM, Truijen J, Kim
YS, et al. Noninvasive continuous arterial blood pressure monitoring with nexfin
®
.
Anesthesiology. (2012) 116(5):1092–103. doi: 10.1097/ALN.0b013e31824f94ed
28. Fischer MO, Avram R, Cârjaliu I, Massetti M, Gérard JL, Hanouz JL, et al. Non-
invasive continuous arterial pressure and cardiac index monitoring with nexfin after
cardiac surgery. Br J Anaesth. (2012) 109(4):514–21. doi: 10.1093/bja/aes215
29. Martina JR, Westerhof BE, Van Goudoever J, De Jonge N, Van Lieshout JJ,
Lahpor JR, et al. Noninvasive blood pressure measurement by the nexfin monitor
during reduced arterial pulsatility: a feasibility study. ASAIO J. (2010) 56(3):221–7.
doi: 10.1097/MAT.0b013e3181d70227
30. Vos JJ, Poterman M, Mooyaart EA, Weening M, Struys MM, Scheeren TW, et al.
Comparison of continuous non-invasive finger arterial pressure monitoring with
conventional intermittent automated arm arterial pressure measurement in patients
under general anaesthesia. Br J Anaesth. (2014) 113(1):67–74. doi: 10.1093/bja/aeu091
31. Imholz BP, Wieling W, van Montfrans GA, Wesseling KH. Fifteen years
experience with finger arterial pressure monitoring: assessment of the technology.
Cardiovasc Res. (1998) 38(3):605–16. doi: 10.1016/S0008-6363(98)00067-4
32. Schutte AE, Huisman HW, van Rooyen JM, Malan NT, Schutte R. Validation of
the finometer device for measurement of blood pressure in black women. J Hum
Hypertens. (2004) 18(2):79–84. doi: 10.1038/sj.jhh.1001639
33. Silke B, McAuley D. Accuracy and precision of blood pressure determination
with the finapres: an overview using re-sampling statistics. J Hum Hypertens. (1998)
12(6):403–9. doi: 10.1038/sj.jhh.1000600
34. Guelen I, Westerhof BE, van der Sar GL, van Montfrans GA, Kiemeneij F,
Wesseling KH, et al. Validation of brachial artery pressure reconstruction from
finger arterial pressure. J Hypertens. (2008) 26(7):1321–7. doi: 10.1097/HJH.
0b013e3282fe1d28
35. Langewouters GJ, Settels JJ, Roelandt R, Wesseling KH. Why use finapres or
portapres rather than intra-arterial or intermittent non-invasive techniques of blood
pressure measurement? J Med Eng Technol. (1998) 22(1):37–43. doi: 10.3109/
03091909809009997
36. Heimark S, Rindal OMH, Seeberg TM, Stepanov A, Boysen ES, Bøtker-
Rasmussen KG, et al. Blood pressure altering method affects correlation with pulse
arrival time. Blood Press Monit. (2022) 27(2):139–46. doi: 10.1097/MBP.
0000000000000577
37. Heimark S, Eitzen I, Vianello I, Bøtker-Rasmussen KG, Mamen A, Hoel Rindal
OM, et al. Blood pressure response and pulse arrival time during exercise testing in
well-trained individuals. Front Physiol. (2022) 13:863855. doi: 10.3389/fphys.2022.
863855
38. Seeberg TM, Orr JG, Opsahl H, Austad HO, Roed MH, Dalgard SH, et al. A
novel method for continuous, noninvasive, cuff-less measurement of blood pressure:
evaluation in patients with nonalcoholic fatty liver disease. IEEE Trans Biomed Eng.
(2017) 64(7):1469–78. doi: 10.1109/TBME.2016.2606538
39. Heimark S, Bøtker-Rasmussen KG, Stepanov A, Haga Ø G, Gonzalez V, Seeberg
TM, et al. Accuracy of non-invasive cuffless blood pressure in the intensive care unit:
promises and challenges. Front Med (Lausanne). (2023) 10:1154041. doi: 10.3389/
fmed.2023.1154041
40. Fleischhauer V, Feldheiser A, Zaunseder S. Beat-to-Beat blood pressure
estimation by photoplethysmography and its interpretation. Sensors. (2022) 22
(18):7037. doi: 10.3390/s22187037
41. Finnegan E, Davidson S, Harford M, Watkinson P, Tarassenko L, Villarroel M.
Features from the photoplethysmogram and the electrocardiogram for estimating
changes in blood pressure. Sci Rep. (2023) 13(1):986. doi: 10.1038/s41598-022-
27170-2
42. Hosanee M, Chan G, Welykholowa K, Cooper R, Kyriacou PA, Zheng D, et al.
Cuffless single-site photoplethysmography for blood pressure monitoring. J Clin Med.
(2020) 9(3):723. doi: 10.3390/jcm9030723
43. Jung Y, Hu J. A K-fold averaging cross-validation procedure. J Nonparametr
Stat. (2015) 27(2):167–79. doi: 10.1080/10485252.2015.1010532
44. Burman P. Estimation of optimal transformations using v-fold cross validation
and repeated learning-testing methods. SankhyāIndian J Stat Ser A (1961–2002).
(1990) 52(3):314–45.
45. Stone M. Cross-validatory choice and assessment of statistical predictions.
J Royal Stat Soc Ser B Methodol. (1974) 36(2):111–47. doi: 10.1111/j.2517-6161.
1974.tb00994.x
46. Hastie T, Tibshirani R, Friedman J. Elements of Statistical Learning: Data
Mining, Inference, and Prediction. 2nd ed New York: New York: Springer
(2009). p. xxii.
47. Myles PS, Cui J. Using the Bland-Altman method to measure agreement with
repeated measures. Br J Anaesth. (2007) 99(3):309–11. doi: 10.1093/bja/aem214
48. Pellaton C, Vybornova A, Fallet S, Marques L, Grossenbacher O, De Marco B,
et al. Accuracy testing of a new optical device for noninvasive estimation of systolic
and diastolic blood pressure compared to intra-arterial measurements. Blood Press
Monit. (2020) 25(2):105–9. doi: 10.1097/MBP.0000000000000421
49. Almeida TP, Cortés M, Perruchoud D, Alexandre J, Vermare P, Sola J, et al.
Aktiia cuffless blood pressure monitor yields equivalent daytime blood pressure
measurements compared to a 24-h ambulatory blood pressure monitor: preliminary
results from a prospective single-center study. Hypertens Res. (2023) 46:1456–61.
doi: 10.1038/s41440-023-01258-2
50. Socrates T, Krisai P, Vischer AS, Meienberg A, Mayr M, Burkard T. Improved
agreement and diagnostic accuracy of a cuffless 24-h blood pressure measurement
device in clinical practice. Sci Rep. (2021) 11(1):1143. doi: 10.1038/s41598-020-
80905-x
51. McGillion MH, Dvirnik N, Yang S, Belley-Côté E, Lamy A, Whitlock R, et al.
Continuous noninvasive remote automated blood pressure monitoring with novel
wearable technology: a preliminary validation study. JMIR Mhealth Uhealth. (2022)
10(2):e24916. doi: 10.2196/24916
52. Tan I, Gnanenthiran SR, Chan J, Kyriakoulis KG, Schlaich MP, Rodgers A, et al.
Evaluation of the ability of a commercially available cuffless wearable device to track
blood pressure changes. J Hypertens. (2023) 41:1003–10. doi: 10.1097/hjh.
0000000000003428
53. Stergiou GS, Palatini P, Asmar R, Ioannidis JP, Kollias A, Lacy P, et al.
Recommendations and practical guidance for performing and reporting validation
studies according to the universal standard for the validation of blood pressure
measuring devices by the association for the advancement of medical
instrumentation/European society of hypertension/international organization for
standardization (AAMI/ESH/ISO). J Hypertens. (2019) 37(3):459–66. doi: 10.1097/
HJH.0000000000002039
54. Lindhard J. Untersuchungen über statische muskelarbeit1. Skand Arch Physiol.
(1920) 40(2):145–95. doi: 10.1111/j.1748-1716.1920.tb01418.x
55. Palatini P. Blood pressure behaviour during physical activity. Sports Med. (1988)
5(6):353–74. doi: 10.2165/00007256-198805060-00002
56. Coulson J, Cockcroft J. Sympathetic influence on the haemodynamic response to
isometric forearm contraction: pP.32.263. J Hypertens. (2010) 28:e528–e9. doi: 10.
1097/01.hjh.0000379801.61894.ba
57. Hartog R, Bolignano D, Sijbrands E, Pucci G, Mattace-Raso F. Short-term
vascular hemodynamic responses to isometric exercise in young adults and in the
elderly. Clin Interv Aging. (2018) 13:509–14. doi: 10.2147/CIA.S151984
58. Brod J, Cachovan M, Bahlmann J, Bauer GE, Celsen B, Sippel R, et al.
Haemodynamic response to an acute emotional stress (mental arithmetic) with
special reference to the venous side. Aust N Z J Med. (1976) 6(suppl 2):19–25.
doi: 10.1111/j.1445-5994.1976.tb03319.x
59. Al’Absi M, Bongard S, Buchanan T, Pincomb GA, Licinio J, Lovallo WR.
Cardiovascular and neuroendocrine adjustment to public speaking and mental
arithmetic stressors. Psychophysiology. (1997) 34(3):266–75. doi: 10.1111/j.1469-
8986.1997.tb02397.x
60. Lamotte G, Boes CJ, Low PA, Coon EA. The expanding role of the cold pressor
test: a brief history. Clin Auton Res. (2021) 31(2):153–5. doi: 10.1007/s10286-021-
00796-4
61. Zhao Q, Bazzano LA, Cao J, Li J, Chen J, Huang J, et al. Reproducibility of blood
pressure response to the cold pressor test: the GenSalt study. Am J Epidemiol. (2012)
176(Suppl 7):S91–8. doi: 10.1093/aje/kws294
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 12 frontiersin.org
62. Khayat R, Kim J, Lee E, Chaudhari A. 1131: a novel sensor to track transient
blood pressure changes during sleep. Sleep. (2024) 47(Supplement_1):A485–A.
doi: 10.1093/sleep/zsae067.01131
63. Wibmer T, Doering K, Kropf-Sanchen C, Rüdiger S, Blanta I, Stoiber KM, et al.
Pulse transit time and blood pressure during cardiopulmonary exercise tests. Physiol
Res. (2014) 63(3):287–96. doi: 10.33549/physiolres.932581
64. Frank SM, Raja SN. Reflex cutaneous vasoconstriction during cold pressor test is
mediated through alpha-adrenoceptors. Clin Auton Res. (1994) 4(5):257–61. doi: 10.
1007/BF01827431
65. ADInstruments. Human NIBP Nano Non-Invasive Hemodynamics Owner’s
Guide. Sydney: ADInstruments (2022).
66. Pour-Ghaz I, Manolukas T, Foray N, Raja J, Rawal A, Ibebuogu UN, et al.
Accuracy of non-invasive and minimally invasive hemodynamic monitoring: where
do we stand? Ann Transl Med. (2019) 7(17):421. doi: 10.21037/atm.2019.07.06
67. Waldron M, David Patterson S, Jeffries O. Inter-day reliability of finapres (
®
)
cardiovascular measurements during rest and exercise. Sports Med Int Open. (2018)
2(1):E9–e15. doi: 10.1055/s-0043-122081
68.HassellundSS,FlaaA,SandvikL,KjeldsenSE,RostrupM.Long-term
stabili ty of cardiovascular and catecholamine re sponses to stress tests: an 18-year
follow-up study. Hypertension. (2010) 55(1):131–6. doi: 10.1161/
HYPERTENSIONAHA.109.143164
69.KimSH,LilotM,SidhuKS,RinehartJ,YuZ,CanalesC,etal.Accuracy
and precision of continuous noninvasive arterial pressure monitoring
compare d with invasive arterial pressure: a systematic review and meta-
analysis. Anesthesiology. (2014) 120(5):1080–97. doi: 10.1097/ALN.
0000000000000226
70. Stokes DN, Clutton-Brock T, Patil C, Thompson JM, Hutton P. Comparison of
invasive and non-invasive measurements of continuous arterial pressure using the
finapres. Br J Anaesth. (1991) 67(1):26–35. doi: 10.1093/bja/67.1.26
71. Wilkes M P, Bennett A, Hall P, Lewis M, Clutton-Broc k TH. Compari son
of invasive and non-invasive measurement of continuous arterial pressure
using the finapres in patients undergoing spinal anaesthesia for lower
segment caesarean section. Br J Anaesth. (1994) 73(6):738–43.doi:10.1093/
bja/73.6.738
Hove et al. 10.3389/fmedt.2024.1464473
Frontiers in Medical Technology 13 frontiersin.org
