See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/329568798

Wireless Health Monitoring System

Conference Paper · December 2018

DOI: 10.1109/BICAME45512.2018.1570510420

CITATIONS READS

7 7,023

4 authors:

Wordh Ul Hasan Mohammad Sultan Khaja

North Dakota State University North South University
21 PUBLICATIONS 61 CITATIONS 2 PUBLICATIONS 13 CITATIONS

SEE PROFILE SEE PROFILE

Saif Ahmed Mohammad Monirujjaman Khan

North South University North South University
2 PUBLICATIONS 13 CITATIONS 320 PUBLICATIONS 3,251 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Wordh Ul Hasan on 07 May 2020.

The user has requested enhancement of the downloaded file.

 Wireless Health Monitoring System
Wordh Ul Hasan, Mohammad Sultan Khaja, Saif Ahmed & Mohammad Monirujjaman Khan
Department of Electrical & Computer Engineering,
North South University, Bashundhara, Dhaka, Bangladesh
wordh.u.hasan@ieee.org, mohammads.k.92@gmail.com, saif.ahmed20@northsouth.edu, monirujjaman.khan@northsouth.edu

Abstract— In Bangladesh many rural areas have as little as

one clinic to help the people. This project aims to help those II. BACKGROUND OVERVIEW
with chronic diseases that prolong with aging. This approach is There are numerous models and with respect to remote
made by taking basic determinants such as SPO2, ECG, patient monitoring and data observing framework. An early
NIBP & temperatures wirelessly transfer them to an online work on this system has been proposed in [3]. The proposed
server. These data can be observed by doctors/caretakers at
framework used to screen the ECG signs of a patient in view
real time from any location being connected online. The
samples of data are acted upon to determine
of Session Initiation Protocol (SIP) and a ZigBee
irregularities of the health conditions of the patient in rest arrange. The framework comprises of a remote ECG sensor,
which are stored on the server every second. Doctors have ZigBee module, SIP Register, a proxy web server, a database
access to the server through a web application which enables server and other wireless devices. Another remote
them to analyses the data. A wireless health monitoring system patient monitoring system based on power line
has been developed to solve this problem, for which an communication (PLC) and the ZigBee network has been
Android and a Web Application was implemented. proposed in [4]. The framework comprises of
physiological sensors like a ZigBee/PLC portal and
Keywords—Patient monitoring, IOT, Healthcare, Wireless some software. The physiological information are gathered
Monitoring, Cardiovascular Disease. by the physiological sensor and are passed to a central server
through a ZigBee/PLC passage. The information are put
I. INTRODUCTION away and broke down at the central server.
The rising epidemic that is known as cardiovascular Based on the ZigBee technology, many patients can be
disease (CVD) in developing countries is creating a monitored simultaneously using physiological signals, which
major challenge for the health sector. From studies found in have been discussed in [5]. The network load is controlled by
[1] & [2] we have observed that there were 1.78 lakh an optimized source routing protocol. Also, in this work
deaths were caused by strokes, 1.06 lakh by ischemic heart wireless network and power related problem, which includes
disease and 28,000 by hypertensive heart disease in 2013 energy consumption, network lifetime delivery ratio, etc.
alone in Bangladesh. have been improved. In [6], a patient monitoring system has
been developed with a low power microcontroller.
Currently available devices are overpriced in the As bedsores are a common problem in hospitals, a resistive
market, even more so, they are not available in Bangladesh. bend sensor is used for it. The system that is proposed can
The resolution was establish a remote personalized assistive support continuous monitoring by enabling mobility of
system which is in need at this moment in this country as patients location and has the potential to replace wired
there are an insubstantial amount of doctors in remote/rural system that is available currently
areas. This system can help those areas greatly and doctors
can work remotely taking better care for many patients who The study in [7] works on a local area network or in a
are in need of assistance. On the other hand, ambulances in personal basis for wireless data acquisition system. The
Bangladesh is are not properly equipped with any sort system uses IEEE 802.15 standard for
of devices that monitor the health of the patient being wireless communication along with Bluetooth technology.
transported, whereas it is a necessity to have a health Monitoring patient using the internet is possible because
monitoring system in ambulances as it gives the doctor of the home based monitoring. The remote patient
critical insight about the patient. Hence, this system can be monitoring system in [8] uses Internet of Things (IoT) and
put to good use within ambulances as well. On the other ZigBee protocol. The real time system can send a report to
hand Bangladesh faces a lot of congestion and traffic on the remote location once analyzing the data
road and there is no emergency lane for the ambulances to
The system developed in [9] wirelessly takes health
pass in critical situations. Many patients face serious
parameters, which are temperature and pulse of a patient. It
consequences for this reason only. So this device can help in
is based on ZigBee and has a good accuracy. On the other
that area severely.
hand, the improved system in [10] not only has portable
The issue addressed in our study makes use of sensors, ECG to monitor patient’s health condition, but also
mobile and web technology to develop a more robust remote enables data collection of patient’s real-time walking
monitoring system catered to heart patients. The solution acceleration. The system can detect falling of patients after
proposed here engages with real-time monitoring analysis of data.
with sensors that are wearable and generate diagnostic
Existing system even though working with some using
information on the fly. These diagnostics are then sent an
web server application or mobile application to monitor
android smartphone via Bluetooth technology and then
the health information of patients they are mostly prototypes
wirelessly transmitted by 3G/Wi-Fi. The patient’s
and uses low performing sensors. Some of them operate
conditions can be checked at any moment of given time by
using a platform and device dependent monitoring system
nearby relatives, caretaker or a doctor.
based on central controller. Hence, overall cost
increases and portability reduces for the whole system.
Moreover in the developing countries, the system used shares the unique key to another person who is the caretaker
to monitor patients has no portability. Hence, only when of the patient then the caretaker or relative can also be
the patient is on bed, it can be used. Other available systems an assistance for the patient to monitor his or her conditions.
are huge, which also make it not portable and costly too and The entity relationship of the system is shown in figure 2.
are only available in ICU of hospitals. On the other hand, Here the three entity are doctor, assistance which is the
other monitoring systems fail to communicate with the patient’s relatives or nurses and the patient himself. The
doctor, which might cause serious situations as they are patient has ECG, SPO2, NIBP and Temp as parameters to be
based on simple circuitry and low level controllers. The monitored. The doctor can take decisions observing these
proposed system we used is very portable and can be carried parameters real-time.
to one place to another and very easily usable in ambulances.
The system provides real time monitoring with reliable
communication over the internet for both doctors and
patient’s caretakers. In terms of network architecture it uses
very reliable Bluetooth communication between sensors and
android device and 4G/WIFI for server communication.
Most of the systems described above uses MySQL databases
where we used noSQL proper for real-time monitoring and
much faster than the available system

III. METHODOLOGY
The system we have developed takes NIBP,
Body Temperature, ECG and SPO2 values through the
sensors attached with the patient. The Sensors are connected
with the module PM6750. The Module sends the data via
Bluetooth to an Android app that we have developed
using the SDK provided by SH Berry Med. The android Figure 2: Entity Relationship Diagram
app needs to be registered by the user and logged in to use The android application was developed using android
the device. After login with the app it searches for the studio and java programming language. We developed
PM6750 module. When connected the app instantly start the API to process the data received by patient
receiving the data and visualize them on the android app monitoring sensors. First we received the data we sorted
also in real time the android app is connected to the server. the data as separation, format conversion and calculation to
We used firebase server for our system for both login display the parameters and waveforms. The system software
authentication and database storage. The firebase server is architecture is shown in figure 3
updated with the data sent via android app over the internet.
The web app developed for this system can be used by the
caretaker or the doctor to monitor the patient’s condition
any given time. The web app pulls the data from the
database. The flowchart has been shown in figure 1. The
details of the system architecture, data flow and components
are explained as follow.

Figure 1: Flowchart of the system

A. System Architecture
There are two entities primarily. Doctor and Patient.
A unique key by the patient shared with the doctor the doctor
can add that patient under his observing conditions. One or
more patients can be observed by one doctor. If the patient Figure 3: System Software Architecture
 The android application receives the data in the network C. Handheld Patient Monitor
layer using Bluetooth communication. In this network layer
we also have checksum to check if the received data is
corrupted or valid. The data package then forwarded to data
packet bus.
In Data processing layer data is received by the data
parser we developed which filters out noise and forwards
valid data for parsing and de-multiplexing into ECG
Package, SPO2 Package, NIBP Package and Temp Package.
These packages then sorts the data in waveforms and
parameters to display. All of these data then put into context
bus. All the data are kept in a buffer queue with maximum
capacity of 256. Figure 5: Handheld Patient Monitor
The Application layer runs drawRunnable class to get the
waveforms from context bus and display in the android For the Patient monitoring module a Multi Parameter
application. It also displays all the other parameters to the Module PM 6750 Patient Monitor System was used that was
android application. While displaying the processed data the developed by Shanghai Berry Electronics. This Bluetooth
application layers also forwards this data to firebase server patient monitor OEM module includes four parameters ECG,
database where web apps can receive the data and display. NIBP, SPO2 and Temperature and can be used with the said
operating systems such as Windows, Linux etc. For our
The web application and server does not process any data project we are using the device with Bluetooth 2.0 with
it only receives the data real-time and display it. The decision android application. An electric oscillator is used for
maker of those data and interpretation is basically the doctor measurement of NIBP and has a cut off time of less than 15
viewing data from the web portal. seconds. With a bandwidth of 0.05Hz-100Hz (3dB), the
ECG has total 6 leads I, II, III, aVL, aVR, aVF
B. System Communication and V1,V2,V3,V4,V5,V6. Carbon dioxide at end of
The system will have 3 kinds of users. The main and respiration is used to measure respiratory rate and
prior users are patients which are connected with the sensors. temperature is measured from vascular or surface
The sensors will send data over Bluetooth the android app temperature. On the other hand, scanning speed is 50mm/s
and the android app will parse those data into a server and for the diagnostic mode 1=40Hz monitor mode and 0.1-25Hz
store them and visualize them. The data is stored in the Operation mode is used and the heart rate is monitored
firebase database. The data is sent from android to the 250 times/minutes. Measurement scope for SPO2 is 35%-
database using firebase user authentication and uploading the 100%.
data via internet to the firebase database under the user’s
individual tables. D. Database Design of the System
For this application we have designed the database not
with traditional relational database but with noSQL
database. Previously seen all the other implementations that
we have seen designed database is using MySQL database
but for this project we implemented with Firebase Database
with noSQL Model. The reason we used noSQL because it
can address several issues that is not easily solved using
relational database. As Patient monitoring system has lots
of data volumes rapidly changing unstructured data. noSQL
is better to address these issues.

noSQL database is best for real-time monitoring

specially in a project like this where we have to deal with
Figure 4: Wireless Communication System lots of data very fast. According to [11] NoSQL is designed
to address:
The Stored data can be viewed real time by the assistive
persons and the doctors. They can view these data using the  Large volumes of rapidly changing structured, Semi-
web app which pulls the data from the firebase via an Structured and Unstructured data.
Angular web app. The web app uses the internet to connect
and login with the firebase database. Given the valid  Agile sprints, quick scheme iteration and
credentials the data is shown in the browser of the assistive frequent code pushes.
person or doctor. Based on the data doctors can give further  Object-oriented programming that is easy to use and
instructions. The instructions can be given by the web portal flexible
as well. And previously given all the prescription and
instructions are also stored in the database from where it  Geographically distributed scale-out
is visible to the web portal. The system communication is architecture instead of expensive, monolithic
shown in Figure 4. architecture.
 Electrocardiogram and SPO2 with respect to time. The
visualization is done with High Charts Library.

Figure 6: Database of the System

In figure 6 we see the database of the system

Figure 7: User Health Information on the web
designed in noSQL. Under the user we have patients and
under patient with each patient’s unique key all the
In figure 8 it shows the portal of the doctor from where
parameters and history of previous prescription prescribed
he can prescribe medicine and exercise to the patient and it
by the doctor is visible. The prescription is stored with
stores in day to day log. The patient history can be viewed
respect to the date they were prescribed. Another reason
by the doctor and the patient portal as well. The details how
we used firebase database here because it uses data
the data is stored is shown in figure 6 and discussed in
synchronization every time data changes instead of typical
database design of the of the system
HTTP Requests and updates everything within
milliseconds and it uses disk to persist data as even if we
are offline firebase can be responsive and once
reconnected to the internet it receives the changes and
synchronizes with current state
E. Web Interface
The web app is implemented using Angular 5, Node.JS,
Bootstrap, HTML, CSS, Typescript, High Charts
and Firebase. The IDE used was Visual Studios.

The web app has two kinds of login authentication

system. One for the patient and another for the doctor. The
patient or the close relative of the patient who is taking
care of the patient can observe the patients conditions
using the patient login also he can have a doctor account and Figure 8: Portal for the doctor to create the report
login and use the key provided by the patient to view the
patient’s condition. The Doctor can get multiple keys for
multiple patient and monitor them using his profile in the F. Android App
web app. In the web app the data is pulled from the At first the users are prompted to register/login on the
firebase database real time and displayed in this table app so they could be indexed as a patient/doctor on the
manner. The doctor can monitor Heart Rate, High Pressure, platform. When the device is set up on the app receives all
Low Pressure, mean Pressure, Pulse rate, Resp. Rate SPO2 the data from PM6750 Module via Bluetooth and displays
and Body temperature. the information to the user.

If the doctor wants to view individual patient’s condition These data related to patient’s health can be accessed by
he can just click on the patient and view the data shown in both the doctor and the patient or their relatives/attendants.
figure 7 it shows the visualization of the data of However, to ensure security of these information both the
doctor and the patient has to login to the app using their conditions of their patients from any location given that they
email address and password. The doctor can monitor the were connected to Wi-Fi/4G networks. We have
patient’s health condition using the app, however, the app tasted with analogue monitoring systems and also our
also uploads the data to the server so that it could also be developed system. There were no significant delay that we
accessed using the website. could notice however we checked the time delay provided
by the specifications of the system used and it says from
android to firebase and firebase to angular web app it takes
around 200 to 300 milliseconds of time delay. Tested with
other systems we found the Heart Rate accuracy ±2
time/minute and SPO2 accuracy: ±2% in the range of
90%~100% ±3% in the range of 70%-89% for the NIBP
measurement accuracy is static pressure ±3mmHg. In figure
11 we see the testing subject whose data is visible in all the
other figures we used in this paper.

Figure 9: Patient Parameter Reading in android app

Here in figure 9 we can see the ECG, SPO2, Temp and

NIBP values displayed in the android app from a patient’s
readings from the sensors.
Figure 11: Testing the system

One of the major contribution we did in this project what

is missing in all the other projects we have discussed before
is the use of Bluetooth and Wi-Fi instead of Zigbee network.
It helped major way in transmission of data and real-time
monitoring

Technology Data Frequency Sensitivity Transmission Link

Rate Power budget

Zigbee 250 2400MHz -98dBm 8dBm 106dB

kb/s

Bluetooth 1MB/s 2400MHz -85dBm 7dBm 92dB

Table 1: Difference between Zigbee and Bluetooth

In table 01 we can clearly see Bluetooth performs better
for real-time monitoring of data.
Also in comparison with noSQL database with relational
Figure 10: Android app to view patient old history database this system outperforms other systems in handling
prescribed by doctor robustness of data the detail comparison is discussed in
database design chapter of this paper.
Figure 10 shows the patient history and old prescriptions The total system costs only 270 USD that also taken the
by the doctor. consideration that the system is not any prototype but
hardware is industrial grade embedded system designed by
IV. TESTS AND RESULTS the Berry SH in china unlike other systems we have
We were able to test our device on some volunteers and discussed in this paper and the application developed for the
monitored the data to check whether there were system can simultaneously handle thousands of patients
any discrepancies in passing the data from the app to the concurrently without any delay. From [12] we know firebase
server. We kept the test subject in a home and connected the can handle 100,000 simultaneous connections and
sensors. The android app that received the data was in the ~100,000/Second simultaneous responses send from a single
same room and was connected via Wi-Fi. The web portal database. The available patient monitoring system in the
was accessed from 5 kilometer away from the test subject by market doesn’t come with web portals and also very pricy
a doctor. Data was successfully transmitted in real time. ranging from on average 400 to 1000 USD seen on ebay and
Doctors were able to participate and remotely check the alibaba stores.
 V. POTENTIAL APPLICATION AREA name and age on web interface and Patient prescription
There are many potential application areas for this history.
system in Bangladesh. Primarily in Bangladesh Traffic jam ACKNOWLEDGMENT
is very severe and our ambulances does not have any kind of
patient monitoring systems. Many heart patient dies on the We would like to express gratitude towards
way to hospital because they couldn’t get the immediate NSU, Electrical and Computer Engineering Department,
attention they needed. Our approach with this project is to Berry Med for helping us to receive the handheld monitoring
system and providing us with the Software development
install these systems in the ambulances so that for these kind
kit for the module. Our faculty advisor Dr. Mohammad
of patient doctor can immediately monitor patients Monirujjaman Khan for guiding us through this project.
remotely and prescribe necessary actions to save lives.

In Rural areas of Bangladesh healthcare system is not REFERENCES

developed as much. And in many areas there is no Doctors [1] Porimol Palma. "Heart diseases now a plague in
Bangladesh," The Daily Star: March 08, 2015
available fulfilling the patient’s necessity. For those areas if
[2] AKM Monwarul Islam, AKM Mohibullah et all,
we introduce these patient monitoring system doctors from Cardiovascular Disease in Bangladesh: A Review, Bangladesh
anywhere can help those patients in need. We need to have Heart Journal Vol. 31, No. 2, 2016
trained nurses or individuals who will keep these system [3] Bonam Kim, Youngjon Kim, InSung Lee, and Ilsum You, Design
running in those rural areas. and Implementation of a Ubiquitous ECG Monitoring System
Using SIP and the ZigBee Networks, In the proceedings of
the Future Generation Communication and Networking
For many patients check-up can be done from home. (FGCN 2007), December 6-8, 2007, Jeju, Korea, pp. 599-604.
All they will need is the sensor data collector and the app. [4] Xi Xueliang, Tao Cheng, and Fang Xingyuan, A health care System
Patients who needs long time monitoring can be monitored based on PLC and ZiGbee, In proceedings of the International
from home using these system by both the doctor and Conference on Wireless Communication, Networking and
Mobile Computing, September 21-25, 2007, Shanghai, China, pp.
patient’s caretaker. This can help the patient by letting them 3063-3066.
stay in their own home in their comfort zones. [5] Fariborz H., Moghawemi, M., and Mehrkanoon, S, The design of an
intelligent wireless sensor network for ubiquitous healthcare, In
the Proceedings of the International Conference on Intelligent
VI. FUTURE WORKS and Advanced System, November 25-28, 2007, Kualalumpur,
Malaysia, pp. 414-417.
For each patient doctor will be able to set parameter
[6] Manohar, A. and Bhatia, D., Pressure Detection and Wireless
ranges for acceptable values. When the data goes out of the Interfaces for patient bed, In the Proceedings of the IEEE Biomedical
range an automated alert and sms will be sent to the doctor Circuits and Systems Conference, November 20-22, , 2007,
and the caretaker for immediate attention. To avoid false Baltimore, MD, pp. 389-392.
alarms we will implement filtering algorithms to detect valid [7] Ashwin K. Whitchurch, Jose K. Abraham, and Vijay K.Varadan,
Member, IEEE., Design and Development of a wireless remote point-
alarms. of-care patient monitoring system, Region 5 Technical Conference,
2007 IEEE.
VII. CONCLUSION
[8] Jingram Luo, Yulu Chen, Kai Tang, Juwen Luo, Remote monitoring
Expanding rate of chronic diseases in aging populace is information system and its applications based on the Internet
turning into a genuine worry because of absence of adequate Things, in the proceedings of the International Conference
on Future Biomedical Information Engineering, December 13-
facilities and to a great degree high cost. The circumstance 14, 2009, Sanya, pp. 482-485.
is far and away more terrible for the general population [9] Chengcheng Ding, Xiaopei Wu, Zhao Lu, Design and
living in remote areas a long way from medical facilities as Implementation of the ZigBee-based Body Sensor Network
delay in analysis and treatment may lead to death. Diagnosis System, In the Proceedings of the 5th International Wireless
Communciations, Networking, and Mobile Computing, September
within time and treatment can solve these issues in a great 24-26, 2009, Beijing, pp.1-4.
deal. In this proposed system the progressions in remote [10] Tian Lan, and Xiaogiong Li, Gait Analysis via a high resolution
patient monitoring system open up the chance of continuous triaxial accelartion sensor based on ZigBee Technology, In the
human services observing frameworks. The created proceedings of the International Conference on Complex Medical
framework is contained sensor modules, Android Technology, May 25-28, Beijing, 2013, pp. 697-702.
application, and web application. The framework is [11] noSQL Database Explained [Online], Available:
https://www.mongodb.com/nosql-explained, Accessed Apr. 14,
successful in monitoring real time with NIBP, SPO2, Body 2018
Temperature and ECG. The collected information is [12] Realtime Database Limits [Online]. Avaliable:
being transmitted to Android handheld gadget utilizing https://firebase.google.com/docs/database/usage/limits
Bluetooth which is then transmitted to web application. Web
application forms got information to indicate patient status,

View publication stats