2012 International Conference on Open Source Systems and Technologies (ICOSST)

978-1-4673-3097-8/12/$31.00 ©2012 IEEE

Design and Implementation of an Embedded System

for transmitting Human ECG and Web Server for

Emergency Services and Remote Health Monitoring A low cost ECG Signal simulator and its transmitter, to send and store data in electronic

databases, in remote location, to be accessed by authorized personnel when needed.

Sundus Abrar

Department of Communication Systems

NUST SEECS

Islamabad, Pakistan

Umair Shahid Aziz

Department of Communication Systems

NUST SEECS

Islamabad, Pakistan

Fareeha Choudhry

Department of Communication Systems

NUST SEECS

Islamabad, Pakistan

Abdullah Mansoor

Department of Communication Systems

NUST SEECS

Islamabad, Pakistan

Abstract— this paper highlights the design of an embedded

telemedicine system and web server for remote health monitoring

of patients. It enlightens the technique to remotely monitor

patient data. We proposed a real time telemedicine system

utilizing GSM/GPRS protocol for the third world countries with

the help of which, the patients can be monitored from any part of

the world by the doctor via the internet. This system utilizes a set

of software simulators and a DAC as a solution to low cost

testing. We have used Lab View software as the software

simulator for generating the ECG signal and NI equipment’s to

get a real time analog output. That signal is fed to a GSM modem

to transmit data over the internet.

ywords— telemedicin; GPRS; embedded system; ECG signals;

web server

I.INTRODUCTION

Health telematics play a major role in improving the lives of patients, particularly in the weaker sections of the society including disabled, elderly and chronically ill patients. It has also been shown that rising hospital expenses are the main factor for rising costs in patient healthcare. Many patients with non-life-threatening illnesses needing health monitoring do not necessarily require hospitalizations. Mobile health monitoring devices bring potential benefits to both patient and doctor; doctors can focus more on priority tasks by saving time normally spent with consulting chronically ill patients and patients can move about in their environment without having to make extensive trips to the doctor – especially if they reside in a remote location. [3]

Telemedicine systems can be used for monitoring of vital clinical parameters from patients at home, and can be utilized for aftercare not only in remote rural but also in urban areas as well [4]. Many studies have demonstrated the applications and advantages of such systems [4].

The advantages of telemedicine are in the areas like:

Rural areas

Ambulances

Transmitting vital signals from one hospital to

another

Heart disease is the main cause of early disability and

premature death in most countries. In modern medicine, there

are sorts of method to diagnose heart disease, such as

electrocardiogram (ECG), ultrasound, MRI, CT and so on.

Among these methods, ECG diagnosis has the advantages of

convenience and low cost so that it can be used in a wide area.

Moreover, most of the cardiac deaths occur outside the

hospital [1]

However the main problem being faced is that the ECG

signal of the patient is only available to doctors present close

the patient. There is no electronic storage of this data and

patients need to travel large distances to get access to their

preferred doctors. A full system should be developed which

can record the ECG of the patient and transfer it a safer place

where doctors from any part of the world can view it and give

their suggestions.

Devices like the Actiwave range of miniature biomedical

waveform recorders are designed to capture EMG, EEG and

ECG signal in daily livings. Each recorder can be taped to the

skin near the position of electrodes. The signal is recorded in it

and then can be downloaded in the PC and analysed [7].

AliveECG software wirelessly transmits through a

Bluetooth connection over the Alive Heart and Activity

Monitor. It is used to monitor patients ECG and heart rate and

for later use [8].

Amon is a wrist worn multi parameter device. As the

device is worn on the wrist so the signals are not clear. It does

have noise factor in it [9].

2012 International Conference on Open Source Systems and Technologies (ICOSST)

Figure 1. Client-Server Approach to the System

A prolonged type of ECG tracing, which is called Holter monitor, provides the physician a better opportunity to capture any abnormal heartbeats or rhythms. The patient wears ECG electrode patches on his/her chest, and the electrodes are connected by wire leads to a recording device. [1] Many researchers have improved the hardware and software of this Holter. Those improvements include store medium, playback methods, analysis algorithms and transmission approach. But the main problems of them are as follows: (1) high cost; (2) could not transmit the ECG information to the doctors anywhere at any time; (3) high power consumption [6].

In order to solve these problems presented above, we have developed a novel telemedicine system which consists of an embedded device that wirelessly transmits ECG signals of a patient to a web server. As opposed to other ECG monitoring devices this system ensures reliable transfer of ECG signal over the internet. The system will present a low cost solution and will be affordable by everyone. This data will be accessible to the doctor anywhere in the world as long as he has an internet connection and the patient is connected to a cellular network.

II. SURVEY

To demonstrate the importance of this system in practical life, a brief survey was conducted covering various local hospitals. A questionnaire report was compiled after the survey: 60 cardio doctors were asked a series of questions to determine how the proposed system would benefit them:

Average number of patients who came to hospital for

ECG in a day is more than 30.

The age of the people who mostly came for ECG is

50+.

40 % doctors said that almost 80% patients suffer most

due to poor consultation.

Average number of patients suffering due to delay in

checkup are almost 40%

In almost all the cases doctor’s physical presence is

required (Mostly in C.C.U.)

Almost all the doctors said that they do not want an

ECG signal over their home.

Old ECG data is always useful for doctors in all cases

III. SYSTEM GOALS

The goal of the proposed Telemedicine system is to ensure availability of best medical services to all regardless of their or their trusted physician’s location. This service provided over a wireless channel guarantees mobility and a wide area coverage. It efficiently utilizes the power and radio resources in a cost effective and automated manner.

The proposed system creates a finite state machine to establish and maintain the network connection which relieves the user from pain of manually re-establishing of connection in case of network failure. An interactive website hosted on a web server is equipped with a secure database that can be accessed by authorized personnel only. This feature ensures integrity and confidentiality of this sensitive data.

IV. SYSTEM OVERVIEW

The system is developed using the client server architecture model. Fig. 1 shows the basic architecture.

A client end can be interpreted as the combination of the ECG collection equipment, user interface and the microprocessor. The server is composed of two parts: a socket based back end and PHP based front end.

The back end is responsible for capturing data packets sent by the GPRS modem. This data is then handled by the front end: storing in database and displaying it on the website. The system flow is depicted in the Fig. 2 and program flow is shown in Fig. 3.

A. Hardware Architecture

For our purposes, we made use of LabView to generate the ECG signal and a digital to analog converter (NI Elvis) to get a real time output. The hardware architecture of the system is shown in Fig. 4.

1) ECG Generation Unit: We came across many methods to generate human ECG signals; one is to get actual electrodes, connect them to the human body and measure the ECG. But this type of signal is very weak and contaminated with allot of noise. The only way to make use of this type of ECG is to filter out the signal and then amplify it. This requires additional circuitry and exceeds the scope of our experiment.

An ECG machine solves these problems for us; by directly providing us the purified human ECG signal. But an ECG machine is very hard to get hold of.

2012 International Conference on Open Source Systems and Technologies (ICOSST)

Patient

Medical Equipment/ECG Machine

Main Processing Unit

Data Send Web Server

Data Recieve

Main Processing Unit

Data Display

Doctor

GPRS network, TCP/IP protocol

Patient Monitoring

SystemRemote Medical Server

Figure 2. Flow of System

Start

Welcome Screen

Enter Credentials

Establish Wireless Connection

with Server

Check for

Successful

Connection

Start ECG Data Transfer

Yes

No

Figure 3. Flow of Program

V. ARCHITECTURAL OVERVIEW

The most feasible option left is to simulate the ECG signal

using specialized software programs. This presents a cheaper

alternative; hence we used LabView 8.6 as platform to

simulate signals as shown in Fig. 5. This software allows us to

generate ECG signal with known strength and frequency on a

PC.

Figure 4. Hardware Architecture

For testing purpose we need the signal in real time. Hence we used a NI ELVIS II (a digital to analog converter created by NI Equipment) to generate ECG signal in real form.

Figure 5. ECG Simulation

2) Control System: A lot of microcontrollers have been used in ECG monitors; from 8-bit to 32-bit as well as quite a few DSPs [5]. In this design, we chose a RISC based microcontroller called ATmegaxx, which is produced by Atmel. It is an 8bit microprocessor with 16bits of in system, self-programmable flash memory and consumes low power. It made the hardware design easy, stable and low cost. The clinical bandwidth used for recording the standard 12-lead ECG is 0.05 – 100 Hz [10] and the standard bandwidth for GSM/GPRS is 900/1800 MHz. Therefore, the microcontroller is set to operate at 64 MHz. This high speed allows capturing maximum of ECG samples for further processing

3) GSM/GPRS Modem: To make the system portable, we have decided to opt for wireless data transmission. Now, there are several wireless technologies that can be used to transmit ECG signals, such as, GSM/GPRS, Bluetooth, ZigBee, WLAN IEEE 802.11 and many others. They have their own characteristics (Bandwidth, Latency, Availability, Security, Ubiquity and so on) [1]. Among all these technologies, GSM/GPRS provides the widest mobility range, covering the largest geographical range.

Hence we chose to work with SimCom’s GSM modem

(Sim300Z) which can be operated to work as a GPRS modem.

It has built-in TCP/IP protocol stack to be the

transmitter/receiver so that the user could send his/her ECG

signals at any time wherever GSM network coverage is

available. Its communication port is UART and it can be used

as a modem in a computer system to connect to Internet. No

dial up connection is necessary and it supports theoretical

speeds of up to 170kbps.

2012 International Conference on Open Source Systems and Technologies (ICOSST)

Fig. 6 demonstrates the various steps involved in

establishing a GPRS connection:

Figure 6. GPRS Connectivity

4) Server Implementation: We designed an actual Web

Server, running live for this experiment. Linux was chosen as

the underlying operating system as it is the leading industry

preference and cost effective as compared to other OS’s. Our

server consists of two ends:

Front End

Back End

a) Back End

Back end of the server is designed to receive the data from

the GPRS. The system in enabled to receive C-sockets that

extract data (ECG Signal) and store it in a secure database.

b) Front End

Front end of the server is used to display the ECG signal

on the internet to the patients and the doctors. PHP and MySql

are used to develop the front end. The digital data received

from the GPRS on the server is reconstructed at the server side

and Google Chart API is used to display the ECG signal.

Doctors and Patients can register themselves and then login at

the website to view the ECG record.

B. Software Architecture

The software is implemented using a layered approach. It

is divided into three major layers as shown in Fig. 7:

Application interface

Software layer and

Hardware Layer

A. Signal Generation

LabView simulates the required electrocardiogram which is

received as real time output at ELVIS. This analogue output is

sent to the microprocessor for analogue to digital conversion.

Figure 7. Layered Architecture

VI. FUNCTIONALITY

B. Transmission

The digital ECG signal from the microprocessor is fed to

the modem that triggers GPRS connectivity as soon as it starts

receiving data. A finite state machine approach is followed for

connection establishment: the microprocessor is programmed

to check for connection establishment and when it detects that

GPRS connection has been compromised, it attempts to re-

establish the connection automatically.

After successful connection establishment the data is

forwarded to the GSM module that converts them into

appropriate data packets by attaching the necessary header

fields. Once this is done, the modem is ready for data

transmission to the web server.

C. Reception at Server

MySQL database is used to keep record of the patients and

the doctors affiliated with the health monitoring program. The

database was designed keeping in focus the following major

entities in the system:

Patient

Doctor

ECG Data

Doctor_ID

1) Description of Entities

Patient: This entity class handles all clients using the

system for their ECG monitoring. The members of this

class possess the embedded hardware and are present at

the client side in the system architecture. The main

attributes of this class are: First_Name, Last_Name,

Address, City, Country, Contact_Num, User_Name,

Password, Serial_Num, Assigned_Doctor_ID

Doctor: This entity class belongs to the clients using the

system to monitor the ECG. These members log in to the

website with their set accounts and view the ECG signal

Idle

Set PDP context

Activate PDP

connection

Establish wireless

connection

Data Transfer

2012 International Conference on Open Source Systems and Technologies (ICOSST)

of their individual patients. They are present at the server

end of the system architecture. Main attributes of this

class are: First_Name, Last_Name, Address, City,

Country, Contact_Num, User_Name, Password,

Designation

ECG Data: This class contains the ECG record of

patients. Its main attributes are: Serial_Num,

Data_Value, Time_Reception

Doctor_ID: This entity class contains the list of patients

assigned to one signal doctor. It is generated at run time

and it is named according to a specific doctor’s ID. For

example, for a doctor who has been assigned the ID

134255, this table would be created as Doctor_134255.

Main attributes of this class are: Serial_Number_Patient,

Disease_Name, Disease_Diagnosis, Check_Up_Date,

Next_Appointment

The entity relationship is shown in Fig. 8.

Figure 8. Entity Relationship Diagram

2) The User Interface: Fig. 9 represents the flow of

information from one end to another on the web page.

On the website, a welcome screen is displayed. The user is

presented with a set of options that allow him to verify if he is

a registered patient, a registered doctor or to simply gain

information about the system and some contact information.

This signal was converted to a real time output at ELVIS.

Home Page

View InfoLogin

View Record Logout

Doctor RegistrationPatient Registration

Contact UsAbout UsRegister

Submit Submit

Patient Login Doctor Login

Is Patient? Patient Login

Is Doctor?

View Patient Search Patient View Record Logout

Doctor Login

Yes

Yes

NoNo

Figure 9: Flow of Website

VII. RESULTS

LabView was integrated with ELVIS II to generate the

ECG signal as shown in Fig. 10.

Figure 10. ECG Signal on LabView

This real time signal was then fed to the microprocessor to

get a digital output (Fig. 11)

The digital data sent to the server via GPRS was received

and stored in the database. This stored data was later used to

reconstruct ECG upon doctor’s request as shown in Fig. 12.

2012 International Conference on Open Source Systems and Technologies (ICOSST)

Figure 11. Digital Data

Experiments showed that the efficiency of the system

increases by increasing the sampling rate at the

microcontroller. This basically increased the number of

samples used to reconstruct the ECG (Fig. 13).

Figure 12. Reconstructed ECG Signal at the Server

Figure 13. Reconstructed ECG after increased sampling rate

VIII. CONCLUSION

In modern day healthcare services, where the health

authorities tend to optimize the resources most effectively, it is

in many cases an advantage to treat/monitor as many patients

as possible at their home.

With this project developed locally, we can ensure

appropriate health care and reliable diagnosis by competent

doctors available to all and sundry regardless of their location

and accessibility. The product developed will be cheaper as

compared to those in the market and will provide as a

prototype for further advancement. We can further replace the

main medical unit with other medical instruments with the

same specifications for future work.

IX. FUTURE DEVELOPMENT

Health Level 7 (HL 7) is a global standard for

communication of patient data between health institutions. In

future we will be working on making this system HL7

compliant. The data packets sent over GPRS will be first

converted into HL7 V3 messages at the client side. The

receiving end shall be an HL7 compliant medical server.

ACKNOWLEDGEMENT

The authors would like to thank their families and friends

who showed great appreciation of the work presented in this

paper.

Especial thanks to Holy Family Hospital, Rawalpindi,

Pakistan for providing us with valuable information regarding

most common and chronic diseases, to the group of doctors

who took out time from their busy schedules and were a great

help during data collection, to DG NUST SEECS, Dr. Arshad

Ali for providing us with equipment and laboratory facilities

to conduct our experiments and all the faculty and staff for

providing every type of support.

Thanks to Dr. Osman Hassan, EE Dept., NUST SEECS for

his support and guidance with the hardware used in the

prototype.

Finally, especial thanks to the scholars and scientists

whose previous research has been very helpful in the

completion of this project.

REFERENCES

[1]. J. Dong, S. Zhang, X. Jia, “A portable intelligent ecg monitor based on wireless internet and embedded system technology” International Conference on BioMedical Engineering and Informatics, 2008.

[2]. Y. Jasemian, “Security and privacy in a wireless remote medical system for home healthcare purpose,” Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology, Aalborg University, Denmark

[3]. P. Chan, Ray and N. Parameswaran, “Mobile e-Health monitoring: an agent-based Approach”

[4]. Yousef, A. N. Lars, “Validation of a real-time wireless telemedicine system, using bluetooth protocol and a mobile phone, for remote monitoring patient in medical practice,” Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology, Aalborg University, Aalborg, Denmark

[5]. S.L.Toral, J.M.Quero, E.M.Perez, and L.G.Franquelo, “A microprocessor based system for ECG telemedicine and telecare,” Materials Research Society Symposium – Proceedings, Materials Research Society, USA, 2001, pp.526-529.

[6]. American Heart Association.

[7]. www.camntech.com/cnt_actiwave.htm

[8]. http://www.alivetec.com/products.htm

[9]. “AMON: A Wearable Multiparameter Medical Monitoring and Alert System” U. Anliker, Member IEEE, J.A. Ward, Member IEEE, P. Lukowicz, Member IEEE, G. Tr¨oster, Member IEEE, F. Dolveck, M. Baer, F. Keita, E. Schenker, F. Catarsi, Member IEEE, L. Coluccini, A. Belardinelli, D. Shklarski, M. Alon, E. Hirt, Member IEEE, R. Schmid, and M. Vuskovic.

[10]. www.mit.edu/~ gari/teaching/6.555/LAB/one/ecg.pdf