A Low Cost Fall Detection Device Base on Accelerometer
Sensor and
Notifications for Vertigo and Syncope Patient
Yulastri, Era Madona, Anggara Nasution
and M. Irmansyah
Jurusan Teknik Elektro, Politeknik Negeri Padang, Jl Limau Manih, Kompus Politeknik Neger Padang, 25218, Indonesia
Keywords:
Vertigo, Accelerometer, Notification, SMS, Syncope.
Abstract: In this study we propose a low-cost portable device to monitor the vertigo and syncope patients whether
normal, dizziness or falls. The purpose of this study is design and manufactures a device using an
accelerometer sensor to determining the condition of patients who are at risk of vertigo and syncope. The
patient's condition is detected by accelerometer which read the values of the x, y and z axes to determine
condition normal, dizzy, or fall. This device also uses a real time clock to remind the patient to take medication
three times a day. The device outputs are LCD Oled, Voice, and SMS notification. If the patient condition is
falls an SMS notification will be sent, the dizzy condition vertigo patient will press the panic button first and
then an SMS notification will be sent. The test results of this device, 30 times falls detection, obtained a
sensitivity value for dizziness activity is 83.3% and a sensitivity value for falling activity is 90%. The system
takes an average of 8 seconds after a drop in activity to send SMS notifications to phone number users, and it
also takes an average of 5 to 6 seconds after a dizzy activity to make phone calls to phone number users.
1 INTRODUCTION
The balance system is an important system for human
life. This system enables humans to be aware of their
position in the surrounding space. Balance is an
integrated system, namely the visual, vestibular,
propioseptic and cerebellar systems. Disorders of the
balance system will cause various complaints,
including a spinning sensation which is often called
vertigo. Vertigo is a common complaint described as
a spinning sensation, shakiness, instability (giddiness,
unsteadiness) and dizziness (Sura, 2010),( Gnerre,
2015),( Thompson, 2005) . Vertigo is a real
public
health problem. The patient has difficulty expressing
the onset of symptoms. According to a
survey from
the Department of Epidemiology, Robert
Koch
Institute Germany in the general population in
Berlin
in 2007, the prevalence of vertigo in 1 year was 0.9%,
vertigo due to migraine was 0.89%, for BPPV 1.6%,
vertigo due to Meniere's Disease 0.51% . In
Indonesia, case data in R.S. Dr. Kariadi Semarang
stated vertigo cases were in the 5th rank of most cases
treated in the neurological ward. The effect of vertigo
on a person is sudden loss of nerve function and
experiencing dizziness so that a person can fall and
faint. Not only vertigo, the cause of fainting or
syncope in the medical world can be caused by, low
blood pressure or dilated blood vessels, irregular
heartbeat, hypoglycemia and neurological diseases
(Kidd, 2016). Because the symptoms of vertigo and
syncope
appear suddenly, it is necessary to take an
action in
the form of supervision for someone who
has experienced it, because if it is not done quickly
action can cause more severe symptoms such as
stroke and even death (Müller, 2019), (Moya, 2009).
Supervision is very important to avoid the not
desirable action. The families always accompany and
supervise what patient. So, if something happens it
can be immediately handled and did not have fatal
consequences. However, this is not an easy thing to
do because families also have their own activities.
Anxiety arises when families cannot monitor what is
happening to vertigo patients. To supervise and
monitor vertigo patients whether there are incidents
of dizziness, falls or not, a device is needed to
determine the position of the patient from a short
distance or far away. Several studies related to health
monitoring have been carried out including for
monitoring heart rate (Irmansyah, 2018) and body
temperature (Kalaithasan, 2018) based on IoT.
Research conducted can also make
a diagnosis using
a smartphone (Trivedi, 2017) and website-based
(Hameed, 2016). The sensor used in this study uses a
Yulastri, ., Madona, E., Nasution, A. and Irmansyah, M.
A Low Cost Fall Detection Device Base on Accelerometer Sensor and Notifications for Vertigo and Syncope Patient.
DOI: 10.5220/0010961800003260
In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 1173-1179
ISBN: 978-989-758-615-6; ISSN: 2975-8246
Copyright
c
2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
1173
pulse sensor and a DS18b20 sensor. The results of this
study have an average error of 1%. All research
conducted only focuses on monitoring using internet
network communication which of course adds to
costs, besides that there is no notification either to the
doctor or the patient's family at risk if something
happens to the patient. In this research we built the
small portable device to monitor vertigo and syncope
patient condition for normal, dizzy or fall. Some
methods had been designed and implemented in this
area especially for fall condition (Jefiza, 2017).
The fall detection approaches are categorized into
three classes, namely: (i) wearable sensor based, (ii)
ambience sensor based and (iii) vision based
(Mubashir, 2013). In this study we used a wearable
sensor based approach category, the advantages of
this approach remain in
the cost efficiency of the
device, the installation and arrangement of the design
are also not complicated. Therefore, the device is
relatively easy to operate.
(Hakim, 2017). In addition,
this tool is also equipped with a panic
button and RTC
(Real Time Clock) to remind patients
to take
medication if they are taking regular
medication with
a sound output. The purpose of this
research is to
design and make a device using accelerometer sensor
which is useful for knowing the
condition of vertigo
and syncope patients. Perhaps, this tool can
monitoring in patients with vertigo and syncope
easily while they are have activity especially
moving.
2 RESEARCH METHOD
The method used in this research is making prototype
tools starting from literature studies, system design,
hardware design, software design, hardware testing,
software and analysis of test results. The stages of
research on Microcontroller-Based Monitoring Tools
for Vertigo and Syncope Patients can be seen in
Figure 1.
2.1 System Design
This tool uses an accelerometer sensor to determine
the tilt angle of the sensor placed on the prototype box
so that the condition of the vertigo patient can be
known. The accelerometer sensor is determined to be
3 conditions, namely, fine, dizzy (accidentally
dropped) and falling. When conditions are fine it will
run normally and for this normal condition a push
button is also provided for a panic button that vertigo
sufferers can use if they need help. Furthermore, in
the second condition, namely dizziness, the patient
experiences dizziness and falls accidentally moving
the body because the balance of the body is disturbed
so that the accelerometer sensor will read the
movements of the patient's body who is experiencing
dizziness. When dizzy, DF-Player will first ask if the
patient is okay, if the patient's condition is not good,
the patient can press the push button to ask for help
by sending an SMS notification and activating the
DF-Player. In the last condition, namely a fall, the tilt
sensor will detect the slope of the patient who fell,
whether he fell left, right, backward and forward. The
design of a monitoring device system for people with
vertigo and syncope based on a microcontroller can
be seen in Figure 2.
Figure 1: Research methodology.
Figure 2: Framework of system design.
Furthermore, the microcontroller will process the
condition that the patient has fallen. Then the DF-
Player will ask the patient's condition first whether
the patient is okay, if there is no response from the
patient, the SIM800L GSM Module will send an SMS
2.2 Hardware System
This series consists of a tilt angle detector using the
MPU6050 accelerometer, push button, RTC, DF
Player, GSM SIM800L module, battery connected to
5V and GND pins, Arduino nano functions as a
microcontroller as shown in Figure 3. MPU6050
accelerometer sensor is connected to pin A4 , A5 on
the microcontroller detects the slope of the patient's
position who is placed on the prototype bag when
experiencing a change in position which is divided
into several conditions. To determine the patient's
condition on the accelerometer sensor can be seen in
the table 2.
iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science
1174
Figure 3: The electronic circuit for the fall detection device.
Table 1: Accelero sensor value for fall detection.
Test value
Patient's
Condition Left
and right
Patient's
Condition Front
and back
X Y Z
< -35
-125 to
125
-125
to
125
Good
(0
o
– 50
o
) =Y
(40
o
– 90
o
) =
X
Good
(0
o
– 50
o
) = Z
(40
o
– 90
o
) =
X
(-35)
to
(- 10)
-150 to -
125
125 to
150
-150 to
-
125
125 to
150
Dizzy
(50
o
– 70
o
) = Y
(40
o
– 20
o
) =
X
Dizzy
(50
o
– 70
o
) = Z
(40
o
– 20
o
) =
X
> -10
-150 <
Y >150
-150
< Y
>150
Falldown
(70
o
– 90
o
) =
Y
(20
o
– 0
o
) =
X
Falldown
(70
o
– 90
o
) =
Z
(20
o
– 0
o
) = X
The accelerometer sensor uses I2C connections (SDA
and SCL) where SDA is connected to pin A4 and SCL
to pin A5 of the microcontroller. The battery as a
mobile voltage source for this system is connected to
+ 5V and GND microcontroller and + 4V to the GSM
SIM800L module. To get a voltage of + 4VDC, a DC-
DC converter is needed. The panic button is
connected to pin 9, the reset button is connected to the
RST and GND pins on the microcontroller. Pins 2,
and 3 are used as the input lines for the output data to
send SMS to the SIM800L. RX and TX pins are used
as output data entry points to activate DF-Player. The
device made will be placed on the patient's belt. The
design of the tool box can be seen in Figure 4.
Figure 4: Tool design for the fall detection device.
2.3 Software Design
Flowchart of vertigo patient monitoring tool with
SMS notification can be seen in Figure 5. The process
starts from the initialization of I/O. In Good
condition, LCD will display the time and date, when
the panic push button is pressed the DF Player will
sound for help, the LCD will display the words
"TOLONG" and SIM800L will send an SMS.
In Dizziness condition, LCD will display “???”
and DF Player will ask the patient's condition, if the
panic push button is pressed DF Player will ask for
help and SIM800L will send an SMS, and if the push
button reset is pressed the DF Player will not sound
and SIM800L will not send SMS, the LCD will
display the time and date. In Drop condition, LCD
will display “???” and the DF Player will ask the
patient's condition, for 10 seconds there is no
response, the SMS will be sent. The LCD will display
"FALL" and the DF Player will sound. If the patient
presses the push button reset at the time of the fall, the
SMS will be canceled and the patient is declared fine.
When OK DF Player will remind you to take
medication 3 times a day and the LCD display
"WAKTUNYA MINUM OBAT".
A Low Cost Fall Detection Device Base on Accelerometer Sensor and Notifications for Vertigo and Syncope Patient
1175
Figure 5: Flowchart of the fall detection device.
iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science
1176
2.4 Testing of Control Tools for Vertigo
and Syncope based on
Microcontroller
The next step is testing a monitoring tool for people
with vertigo and syncope based on a microcontroller,
this test aims to determine the advantages and
disadvantages of the system that has been created.
The test is carried out in two stages. First, testing
system performance, second, testing tool
notifications. This bag prototype-shaped device is
then attached to the patient, as seen in Figure 6.
Figure 6: Installation of fall detection device on the body.
3 EXPERIMENT RESULT
3.1 Testing the Performance System
This test is carried out to determine the performance
of the device, whether the patient is in a dizzy state,
normal activities and fall detection.
Figure 7: The implementation and testing of fall detection
of vertigo ang syncope patient.
Starting by testing the tilt sensor on the
accelerometer, we take the output data on the X, Y
and Z axes to detect the state of the patient. The data
taken is based on the slope of the angle 0o, 50o, 70o,
and 90
o back and to the left then the sensor data is
displayed on the serial monitor, so that the results are
in accordance with table 2.Based on table 2 when the
patient's condition is fine the accelerometer sensor is
tilted 0o - 50o on the Y axis and 90o - 40o on axis X, the
sensor value shows -125 s / d 125 on the Y axis, the
sensor value is <-35 on the X axis. For dizzy
conditions the sensor value taken is at a slope of 50
o -
70o at the Y axis and 40o - 20o on the X axis, the sensor
values show -150 to -125 and 125 to 150 on the Y
axis, the sensor value -35 to -10 on the X axis. For
falling conditions the sensor values are taken is at a
slope of 70o - 0o on the Y axis and 20o - 0o on the X
axis, the sensor value shows -150 <Y> 150 on the Y
axis, the sensor value> -10 on the X axis, for the left
and right directions with the rotating point on the Z
axis , and if the direction of the front and back of the
Y axis will be replaced with the Z axis and tit The
rotary angle is on the Y axis.
Changes in the angle value will affect the value of
the accelerator sensor. The greater the angle change,
the greater the value of the accelero sensor. So when
the accelero sensor value shows a value> 150 or
more, it can be said that people with vertigo have
fallen. Then the DF-Plyer will sound and within 10
seconds there is no response from the patient, the
SIM800L will send an SMS. Furthermore, motion
testing is carried out in three positions to detect
sudden dizziness experienced by the patient. The test
results can be seen in table 2.
Table 2: Dizziness detection tests on regular activities.
Position
Numbe
r
of
experi
me
nt
Dizzy
Notifications
via phone call
Accu
racy
%
Detection
%
Yes No Yes No
Sit-Stand 10 8 2 80
83.3
16.6
7
Standing
-Walking
10 8 2 80
Walking-
Sit
10 9 1 90
Total 30 25 5
Table 3: Detection test falls on the system.
Category Number of
experiment
Notification
Fall
Accuracy
%
Detection
%
Yes No
Face down 10 9 1 90 Yes No
Recumbent 10 9 1 90 90 10
Total 20 18 2
For thirty times the trial results testing in Table 2, 83%
of the devices can detect patient dizziness during
normal activities, with an accuracy rate of about 80%
A Low Cost Fall Detection Device Base on Accelerometer Sensor and Notifications for Vertigo and Syncope Patient
1177
to 90%. Furthermore, testing the detection of tools for
falls and prone incidents is carried out. The test
results
can be seen in table 3. Based on the
experiments in
table 3, some falling activities such as falling on your
back and falling on your stomach can be detected by
the system as falling activities with an accuracy rate
of 90%.
3.2 Testing the Notification of Tool
This test is done to find out whether the device can
send SMS notifications and phone calls if the patient
is dizzy and falls. In this tool, the SIM800 GSM
module is used for sending SMS and telephone calls
as shown in Figure 8. Table 4 is the test results of
phone calls conducted ten times.
Based on the experiment in table 4, it takes
SIM800L to make a phone call with a duration of 5 to
6 seconds. Then performed testing of tools for
sending sms. The following is a table of the results of
testing the SMS delivery which was carried out ten
times
Figure 8: Testing Notification Tools for sending sms.
Table 4: Testing SIM800L timeout test to make phone calls.
Experiment Phone call
Time period
(
second
)
1 Yes 5
2 Yes 5
3 Yes 5
4 Yes 6
5 Yes 6
6 Yes
5
7 Yes 5
8 Yes 5
9 Yes 5
10 Yes 6
Table 5: Testing the time period for sending SMS
notifications.
Experiment Phone call
Time period
(second)
1Yes 8
2Yes 9
3Yes 8
4Yes 8
5Yes 7
6Yes 9
7Yes 10
8Yes 8
9Yes 9
10 Yes 8
Based on the experiment in table 5 the time it
takes for SIM800L to send an SMS notification
with a duration of 8 seconds. When SIM800L
makes a
phone call, it cannot send SMS
notifications at the
same time. If done together, the
SIM800L will be an error.
4 CONCLUSIONS
In this study, we made a low-cost portable device
to
monitor and monitor patients at risk of vertigo
whether dizziness, falls or not occurred. The results
of tests carried out on the device to detect falls as
much as 30 times, obtained a sensitivity value for
dizziness activity of 83.3% and a sensitivity value for
falling activity by 90%, meaning that the system is
able to detect dizziness and fall activity quite well. an
average of 8 seconds after the activity falls to send
SMS notifications to phone number users, and it also
takes an average of 5 to 6 seconds after a dizzy
activity to make phone calls to phone number users.
REFERENCES
D. Sura and S. Newell, “Vertigo-diagnosis and management
in the
primary care,” Br J Med Pr., vol. 3, no. 4, p.
a351, 2010, [Online]. Available: http://www.
bjmp.org/content/vertigo- diagnosis-and-management-
primary-care.
P. Gnerre, C. Casati, M. Frualdo, M. Cavalleri, and S.
Guizzetti,
“Management of vertigo: From evidence to
clinical practice,” Ital. J. Med., vol. 9, no. 2, pp.
180–192, 2015, doi:
10.4081/itjm.2015.437.
T. L. Thompson and R. Amedee, “Vertigo: A review of
common peripheral and central vestibular disorders,”
Ochsner J., vol. 9, no. 1, pp. 20–26, 2009.
K. Kidd, C. Doughty, and S. Z. Goldhaber, “Syncope
(Fainting),” Circulation, vol. 133, no. 16, pp. e600–
iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science
1178
e602, 2016, doi:
10.1161/CIRCULATIONAHA.115.017308.
P. Müller-Barna et al., “TeleVertigo: Diagnosing Stroke in
Acute Dizziness: A Telemedicine-Supported
Approach,” Stroke, vol. 50, no. 11, pp. 3293–3298,
2019, doi:
10.1161/STROKEAHA.119.026505.
A. Moya et al., “Guidelines for the diagnosis and
management of syncope (version 2009),” Eur. Heart
J., vol. 30, no. 21, pp. 2631–2671, 2009, doi:
10.1093/eurheartj/ehp298.
M. Irmansyah, et al., “Low Cost Heart Rate Portable
Device for Risk Patients with IoT and Warning
System,” IEEE International Conference on Applied
Information Technology
and Innovation (ICAITI), pp.
46-49, 2018.
M. G. Ayoub, M. N. Farhan, and M. S. Jarjees,
“Streaming in- patient BPM data to the cloud with a
real-time monitoring system,” Telkomnika
(Telecommunication Comput. Electron. Control., vol.
17, no. 6, pp. 3120–3125, 2019, doi:
10.12928/TELKOMNIKA.v17i6.13263.
M. Subito, M. Ikhlayel, and E. Setijadi, “Internet of things-
based vital sign monitoring system,” Int. J. Electr.
Comput. Eng., vol. 10, no. 6, pp. 5891–5898, 2020,
doi:
10.11591/ijece.v10i6.pp5891-5898.
K. Kalaithasan, N. A. M. Radzi, and H. Z. Abidin,
“Internet of things application in monitoring sick
building syndrome,” Indones. J. Electr. Eng. Comput.
Sci., vol. 12, no. 2, pp. 505– 512, 2018, doi:
10.11591/ijeecs.v12.i2.pp505-512.
M. Irmansyah, E. Madona, and A. Nasution, “Design and
application of portable heart rate and weight measuring
tool
for premature baby with microcontroller base,”
Int. J. GEOMATE, vol. 17, no. 61, pp. 195–201,
2019, doi:
10.21660/2019.61.ICEE12.
S. Trivedi and A. N. Cheeran, “Android based health
parameter monitoring,” Proc. 2017 Int. Conf. Intell.
Comput. Control
Syst. ICICCS 2017, vol. 2018-Janua,
pp. 1145–1149, 2018, doi:
10.1109/ICCONS.2017.8250646.
Yuhefizar, A. Nasution, R. Putra, E. Asri, and D. Satria,
“IoT: Heart Rate Monitoring Tool Using Android
with Alert
Messanger Telegram System,” IOP Conf.
Ser. Mater. Sci.
Eng., vol. 846, no. 1, 2020, doi:
10.1088/1757- 899X/846/1/012014.
M. Islam et al., “Android based heart rate monitoring and
automatic notification system,” 5th IEEE Reg. 10
Humanit.
Technol. Conf. 2017, R10-HTC 2017, vol.
2018-Janua, pp. 436–439, 2018, doi: 10.1109/R10-
HTC.2017.8288993.
R. T. Hameed, O. A. Mohamad, O. T. Hamid, and N.
Ţǎpuş, “Patient monitoring system based on e-health
sensors & web services,” Proc. 8th Int. Conf. Electron.
Comput. Artif. Intell. ECAI 2016, 2017, doi:
10.1109/ECAI.2016.7861089.
V. Jones, V. Gay, and P. Leijdekkers, “Body sensor
networks for mobile health monitoring: Experience in
Europe and Australia,” 4th Int. Conf. Digit. Soc. ICDS
2010, Incl.
CYBERLAWS 2010 1st Int. Conf. Tech.
Leg. Asp. e-Society, pp. 204–209, 2010, doi:
10.1109/ICDS.2010.41.
A. Jefiza, E. Pramunanto, H. Boedinoegroho, and M. H.
Purnomo,
“Fall detection based on accelerometer and
gyroscope using back propagation,” Int. Conf. Electr.
Eng. Comput. Sci.
Informatics, vol. 2017-December,
no. September, pp. 19–21, 2017, doi:
10.1109/EECSI.2017.8239149.
Y. Lee, H. Yeh, K. H. Kim, and O. Choi, “A real-time fall
detection system based on the acceleration sensor of
smartphone,” Int. J. Eng. Bus. Manag., vol. 10, pp. 1–8,
2018, doi:
10.1177/1847979017750669.
D. Aphairaj, M. Kitsonti, and T. Thanapornsawan, “Fall
detection system with 3-axis accelerometer,” J. Phys.
Conf. Ser., vol. 1380, no. 1, 2019, doi: 10.1088/1742-
6596/1380/1/012060.
M. Mubashir, L. Shao, and L. Seed, “A survey on fall
detection: Principles and approaches,”
Neurocomputing, vol. 100, pp. 144–152, 2013, doi:
10.1016/j.neucom.2011.09.037.
A. Hakim, M. S. Huq, S. Shanta, and B. S. K. K. Ibrahim,
“Smartphone Based Data Mining for Fall Detection:
Analysis
and Design,” Procedia Comput. Sci., vol.
105, no. December 2016, pp. 46–51, 2017, doi:
10.1016/j.procs.2017.01.188.
A Low Cost Fall Detection Device Base on Accelerometer Sensor and Notifications for Vertigo and Syncope Patient
1179
