KeepSafe®
Wristband Device for Heart-rate Monitoring
Vitor Pinto
1
, Raquel Sousa
1
and Gil Gonçalves
2
1
Increase Time, Matosinhos, Portugal
2
Faculdade de Engenharia da Universidade do Porto, Porto, Portugal
Keywords: Bio-signals Monitoring, Mobile Devices, Photopletysmography, Heart Rate Estimation, Wristlet Monitor,
Mobile Application, Signal Processing.
Abstract: The goal of this project is to meet efficient and technological management solutions that allow to increase
life quality of elderly people and at same time reduce costs in the health sector. The KeepSafe® wristband
device will increase the safety of users since it will be able to continuously monitor vital signs and it will
generate dynamic alerts based on thresholds or manual alerts by pressing a SOS button. The device uses a
reflective PPG (Photoplethysmogram) technique using two green LEDs and a photodiode sensor on the
wrist to continuously estimate heart rate. Then this data is correlated with the activity data estimated by an
accelerometer and gyroscope. All data is sent to a smartphone via Bluetooth® Smart. The future goal is to
make the device autonomous by adding GSM communication capabilities.
1 INTRODUCTION
Nowadays, the ageing of population is one of the
biggest problems of our society. It´s predictable that
the number of elderly people will duplicate by the
middle of century. This demographic change will
cause an increase of health problems, in particular
heart diseases. In Portugal the rate of deaths caused
by heart diseases is above 30% (Instituto Nacional
de Estatística, 2012) and often a life could be saved
with a tele-monitoring system.
Continuously monitoring of vital signs and daily
activity information can help to promote better
healthcare and better quality of life. Clinical data is
obtained by specific equipment, mostly available in
hospitals, with several sensors, being necessary to
put into practice the concept of proximity to be able
to access in real-time health related information
wherever the patient is.
IncreaseTime is a technology-based company
whose activity is centred in ICT and wireless sensors
solutions to promote people’s life quality, especially
focused on patients with chronic diseases and elderly
people.
The KeepSafe® project intends to create a low-
cost and non-invasive device for daily use to
continuously monitor the heart rate and alert a
caregiver in case of an emergency episode. The
device is based on a wristband which is comfortable
for daily use and the wrist is the suitable place to
measure heart rate through the
photoplethysmography method.
Figure 1: KeepSafe wristband monitor.
Our monitoring system has three main
components: the KeepSafe® device, a mobile
application and a web application. The device uses
the sensor to measure the heart rate and sends it to
mobile application through Bluetooth® technology
that processes data locally and warns the caregivers
if the data exceeds the thresholds. Furthermore, the
device has an alert button on the wristband for
emergency situations which immediately establishes
83
Pinto V., Sousa R. and Gonçalves G..
KeepSafe
R
- Wristband Device for Heart-rate Monitoring.
DOI: 10.5220/0005184500830087
In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2015), pages 83-87
ISBN: 978-989-758-071-0
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
the communication between the patient and the
caregiver. This function could have a vital role in
people lives. After process data, the mobile device is
responsible for sending it to a remote server where
caregivers can follow up the health status of users
through a web application. This system takes
advantage from the communication interfaces and
processing power of handheld devices like a
smartphone to create a low cost and user-friendly
monitoring system.
The use of devices for personal health
monitoring systems is an emergent area of research.
Related to this work we can enumerate projects that
have some common goals and projects with a
broader scope. The eCAALYX project (Boulos et
al., 2011) inserted in the AAL programme offers
solutions for prevention and management of chronic
conditions in elderly people. This project is based on
a t-shirt with integrated biosensors and
communication of bio signals into a smartphone
application with a simple and intuitive interface. The
accessible parameters in the application are
respiratory rate, temperature and activity type
(walking, standing or lying).
Other work based on a t-shirt is the VitalJacket
(BioDevices S.A., 2008). This device uses wearable
technology for continuously monitoring heart rate
and ECG waves. The data is saved in the device
attached to the t-shirt and then analysed by a
physician. Another work with identical goals is the
WIHMD (Kang et al., 2006). The wrist-worn
integrated health monitoring device is a multi-
parameter wristlet that includes 5 bio-signals and a
fall detector. The bio-signals measured by WIHMD
are ECG (single lead), blood pressure, SPO2,
respiration rate, and body temperature. This device
communicates with a mobile phone giving to the
system tele-reporting functions to advise a caregiver
in emergency situations. The Oxitone device
(Oxitone Medical Ltd., 2013) is a wristband based
device developed by an Israeli Company to monitor
Heart Rate and oxygen saturation (SPO2). The
Oxitone device uses a PPG sensor to perform these
measures and has the ability of sending this data via
Bluetooth® to a given application. The application
performs data analysis and alerts the treating
physician.
Related to mobile applications there are a lot of
them in the market but the vast majority are focused
on fitness activities and doesn’t have the feature of
monitoring multiple bio-signals. Recently, some of
these applications are changing their focus for health
records like TactioHealth (Tactio Health Group,
2014). This application has connectivity to several
electronic health devices and is used to control a set
of parameters like weight, blood pressure, lifestyle,
heart rate, cholesterol and glucose.
The KeepSafe® system has similarities with the
presented works but we intend to differentiate it by
creating an innovative and complete system that
works autonomously (no other device or application)
and automatically alerts formal or informal
caregivers.
This article contains 3 more sections. In section 2
the method is explained describing the KeepSafe®
device, signal acquisition and required processing to
compute heart rate. Section 3 presents experimental
results and discussion. Finally, section 4 refers
conclusions and future work.
2 METHOD
In this section we will describe in detail the
KeepSafe® device and signal acquisition.
2.1 KeepSafe
®
Device
The device is composed mainly by 5 blocks:
accelerometer/gyroscope, a PPG sensor combined
with an analog front-end (AFE4490, Texas
Instruments, USA) for pulse oximeter applications, a
processor unit, an emergency button and a
Bluetooth® unit.
The selected accelerometer (MPU9150,
InvenSense, USA) has an integrated gyroscope too,
this component allows to perform activity
recognition tasks through patterns analysis. This was
important to correlate heart rate measurement with
activity level and it provides extra information about
user daily life. The accelerometer could also be
helpful to estimate the error in PPG signal. This
information makes possible to use adaptive filtering
techniques to improve signal-to-noise ratio.
The PPG sensor and front-end is one of the
central features once heart rate is the critical factor
to monitor. The analog front-end allows controlling
LEDs (Light-Emitting Diodes) to pulsate light,
periodically, through the wrist. Then the sensor will
capture the reflected light which is quantified by the
front-end. All the data is sent to the mobile
application using the Bluetooth® smart unit. Figure
2 illustrates the PPG sensor with the LEDs placed at
each side. The SOS button is an extra safety for the
user: in case of an emergency episode, users can
press the button and the alert will be sent by
Bluetooth®. All alerts (manual or dynamic) are
forwarded to the caregiver through the mobile
BIODEVICES2015-InternationalConferenceonBiomedicalElectronicsandDevices
84
application. The work is still in progress and the
next step is to add new features like GSM
communication and GPS antenna. These new
features will eliminate the smartphone requirement.
The mobile application (Sousa et al., 2012) is
responsible to collect, process and send data to the
remote server and it allows users to configure
parameters like personal data, thresholds for heart
rate and has an interface to visualize data in real-
time. The mobile application connects to a remote
server developed using the PlugThings® Framework
(FreedomGrow, 2011) which is very useful to
remote monitor users and follow up their health
status based on history monitoring data.
Figure 2: Heart rate sensor in KeepSafe device.
2.2 Bio-signal Acquisition
PPG is one of pulse wave detection methods which
consists in emitting light through vessels or arteries
and measure the amount of light that crosses the
tissue (transmissive PPG) or the amount of light that
reflects back (reflective PPG) (Tamura et al., 2014).
This technique uses as principle the variation in light
absorption when the blood is flowing to calculate the
pulse rate. In this particular case, we use the
reflective technique because it is more suitable for
measurements on the wrist. We use two LEDS for
emitting light (operating in the green wavelength
520nm) and a correspondent wavelength photodiode
sensor. That wavelength is proved to be suitable for
PPG acquisition at wrist. Studies about different
wavelengths in reflective PPG have been done (Lee
et al., 2013). For the signal acquisition we use a
sample rate of 100 Hz once is good enough to get
pulse rate frequency and at the same time doesn’t
overload the processor leaving more processing
power for signal filtering, heart rate compute
algorithm and other tasks.
2.3 Bio-signal Processing
Since PPG signal is a low amplitude and noisy
signal it’s necessary to add extra signal processing in
order to improve signal quality and heart rate
estimation.
The first processing stage begins in the front-end
when the signal passes through a transconductance
amplifier. After that the signal is submitted to a
second stage amplifier and ambient light
cancellation and the final step is to submit the signal
to ADC converter to get a digital form signal. All
that processing is performed in the front-end.
The digital signal is then read from the master
processor and digital filters are applied to get the
desired frequency range and improve signal quality
eliminating some of the noise caused by movements.
The selected processor (ATXMEGA128A4U,
Atmel, USA) is a 16 bit processor which is
characterized by being low-cost and low power
consumption at sleep mode. Besides, that processor
provides enough computing power to deal with all
the tasks and signal processing.
The selected digital filter was a pass-band IIR
butterworth filter. The selected type was IIR since
the FIR type requires a high order filter to get the
desired behaviour which would demand more
processing power. An eight order IIR filter was used
and the selected range frequency was 0.6-4.5 Hz.
That frequency is enough to get pulse rate range
between 36 and 270 bpm without attenuation.
3 RESULTS
In this section it will be presented experimental
results regarding to PPG wave and heart rate
calculation.
The signal showed in figure 3 is an example of
the signal after applying the filters described in
previous section.
Figure 3: PPG signal obtained with KeepSafe device.
Heart rate estimation is calculated by extracting
KeepSafe®-WristbandDeviceforHeart-rateMonitoring
85
information from the filtered wave. We can clearly
identify the wave peaks in the image. The developed
algorithm is based on the peaks detection to
calculate the distance between each peak. In order to
detect peaks were used two adaptive threshold based
on the last two seconds of data. One high threshold
and a low threshold to avoid fake peaks. Each time
the wave crosses high threshold we are facing a
possible peak. At this time the value in x-axis (the
time axis) corresponding to local maximum is
stored. As soon as the wave crosses the low
threshold that peak is confirmed and the algorithm
starts looking for the next peak. Furthermore another
verification condition was to set a minimum distance
of 200 milliseconds between peaks.
The computed heart rate is sent to mobile
application where information is shown as illustrated
in the picture below. All user information can also
be consulted through the Web Application.
Figure 4: Main screen of mobile application.
Heart rate measurements obtained with the
KeepSafe
®
device were evaluated in comparison
with a conventional pulse oximeter (CMS50E,
Contec, China) which has an accuracy of ±2 bpm
according to manufacturer. The results were very
close, showing a deviation of ±2 bpm between both
devices which implies that KeepSafe
®
has a
maximum error of ±4 bpm. Tests were performed in
stationary conditions. It was observed that a
significant movement affects the PPG signal making
harder to get precision measurements. That will be
an aspect to improve in future with adaptive filters
bearing in mind the person movement.
4 CONCLUSIONS
During the project development we conclude that it
is possible to develop a user friendly and low-cost
mobile monitoring system. Reflective PPG at wrist
provides a good method to develop a comfort heart
rate monitor. However, it implies to have efficient
algorithms and a considerable processing power.
There is other similar wrist monitors but with
focus on sports activity. KeepSafe
®
intends to be
mostly used by elderly people and works as an
emergence device to increase user’s safety.
The future work involves the development of an
activity recognition algorithm to correlate this data
with heart rate monitor. This information will be
useful to apply adaptive filtering techniques that
leads to higher signal-to-noise ratio during motion.
The battery autonomy is another aspect to improve.
The use of management power techniques like sleep
mode of peripherals whenever they aren’t operating
will allow extend autonomy to encourage continued
use of KeepSafe® monitoring device.
Regarding hardware, a GSM and GPS modules
will be integrated in order to transform the
KeepSafe® bracelet in an autonomous device.
REFERENCES
BioDevices, S.A. (2009). Vital Jacket. Retrieved 14 July,
2014 from http://www.vitaljacket.com/
Boulos, M. N. K., Wheeler, S., Tavares, C., and Jones, R.
(2011). How smartphones are changing the face of
mobile and participatory healthcare: an overview, with
example from eCAALYX. Biomedical engineering
online, 10 (1):24.
FreedomGrow (2011). PlugThings Framework. Retrieved
14 July, 2014 from http://www.freedomgrow.pt/
Instituto Nacional de Estatística, I. (2012). Indicadores
Sociais 2012. Retrieved from http://www.ine.pt/
xportal/xmain?xpid=INE&xpgid=ine_indicadores&in
dOcorrCod=0004280&contexto=bd&selTab=tab2.
Kang, J., Yoo, T., Kim, H. (2006). A Wrist-Worn
Integrated Health Monitoring Instrument with a Tele-
Reporting Device for Telemedicine and Telecare.
IEEE Transactions on Instrumentation and
Measurement, Vol. 55, No. 5, pages 1655-1661.
Lee, J., Matsumura, K., Yamakoshi, K.-I., Rolf, P.,
Tanaka, S., Yamakoshi, T. (2013). Comparison
Between Red, Green and Blue Light Reflection
Photoplethysmography for Heart Rate Monitoring
During Motion. IEEE Engineering in Medicine and
Biology Society, Vol., no., pages 1724-1727.
Oxitone (2013). Oxitone wrist-watch. Retrieved 14 July,
2014 from http://oxitone.com/
BIODEVICES2015-InternationalConferenceonBiomedicalElectronicsandDevices
86
Sousa, R., Alves, M., and Gonçalves, G. (2012).
Healthcare management with KeepCare.
Tactio Health Group (2014). TactioHealth Application.
Retrieved 11 July, 2014 from http://
www.tactiosoft.com/tactiohealth.
Tamura, T., Maeda, Y., Sekine, M., Yoshida, M. (2014).
Wearable Photoplethysmographic Sensors—Past and
Present. Electronics 3, No. 2, pages 282-302.
KeepSafe®-WristbandDeviceforHeart-rateMonitoring
87