JOURNAL OF TELECOMMUNICATIONS, VOLUME 20, ISSUE 1, MAY 2013 13

Healthcare Monitoring based on Wireless Sensor Network with New Features

Mahmoud Shaker, Samir Jasim and Akram Jaddoa Khalaf Department of Electrical Engineering, University of Babylon Hilla Iraq
Abstract Development of healthcare monitoring systems based on wireless sensor network (WSN) has seen a great increase extensively in previous years and so far. These systems are more prevalent because of the importance of human life and the provision of appropriate techniques to be applied. So they should be suitable for healthcare requirements.In the presented paper we designed a real-time healthcare monitoring system based on WSN. The system senses and displays ECG, Heart Rate, and temperature of a patient. The ECG is the main sensor in this system and the heart rate is calculated programmatically from ECG signal. The temperature of the human body is measured using thermistor. The system is characterized by its friendly GUI, easy to use, and very simple to manage by the doctor in charge. In addition that the system has the ability to detect and put in work any new node (patient node) automatically also when any node fails in operation for any cause, the system detects it and put it out of service automatically. Also in the designed system we add the doctor node; this node is movable (doctor accompined when he is out of Intensive Care Unit, ICU) wich gives him an alarm signal for any ubnormal case. In relation to the network used in this work is star topology and the wireless protocol is ZigBee. The software used is LabVIEW as a processing, graphical user interface (GUI), and an auto alarm. This system operates in hospital healthcare and it can be operates in home healthcare.The presented system is installed in Merjan Hospital Babylon city and tested by the hospital doctors for many cases; the obtained results are approved and very satisfactory.

Index Terms ECG, Healthcare, Heart rate, WSN, ZigBee.

1 INTRODUCTION

he Wireless sensor network (WSN) dominated many applications like Healthcare, environment, industrial, and multimedia and the recent years have high attention in field of healthcare monitoring *1-3+. The healthcare monitoring systems save humans live using alarm system for emergency case, storing the information in database for medical research, and decreasing the cost of medical services *1+. As a result of development of sensors, batteries, processors, and wireless communication systems became possible implementation of wearable device. So WSNs have opportunity to realization the wireless monitoring system. The common most cause of death is the heart disease *4+. So the electrocardiograph (ECG) can identify any case out of the ordinary activity of the heart. The most appropriate area of application for the WSN is ECG monitoring system *4+ .The most important point in these applications is the real-time (no delay). The WSNs features are high reliability, low cost,

Mahmoud Shaker Nasr has Ph.D. degree in electrical Engineering and assistant professor with the department of electrical engineering, university of Babylon, Hilla, Iraq. Samir Jasim has Ph.D. degree in electrical engineering and assistant professor with the department of electrical engineering, university of Babylon, Hilla, Iraq. Akram Jaddoa Khalaf had H.D. Degree and is currently M.Sc. student in the department of electrical engineering, university of Babylon, Hilla, Iraq.

low data rate, low power, and real-time. All these features make the WSN suitable for healthcare applications. ZigBee/IEEE 802.15.4 protocols are developed to allow different systems to work together. WSN requirements are the reason to develop global hardware and software standard (ZigBee/IEEE 802.15.4) *5+. The physical and MAC layers, are defined the 802.15.4 and ZigBee are defined the network and application layers *2+. The WSN operate in the 433 MHz, 868 MHz, 915 MHz, and 2.4 GHz license-free Industrial Scientific Medical (ISM) frequency bands *1,6+. The ZigBee has many network topologies (star, tree, and mesh). Each method has its advantages that make it suitable for a particular application. Star topology does not contain any delay and easy to built, but the network will be limited coverage. Either tree topology makes it scalable, but contains delay and complex. Where the mesh topology has same advantages of the tree topology moreover it has more robustness and more complexity. The core of WSN is the nodes which can be coordinator, routers, or end devices. The coordinator is the heart of the network which connects the other nodes. Depend on the network topology the end nodes connect to router nodes or directly to coordinator. In this paper, we design a real-time monitoring system for multi-patient and auto alarming system. The ECG, heart rate, and temperature for each patient are collected and sent it to the computer for monitoring and alarm

14

system. In case of any emergency for the patient the system automatically send an alarm for doctor node.

2 SYSTEM ARCHITUCTURE
We developed the healthcare system is shown in the block diagram figure (1).

Base station GUI Gateway Processing Alarm

Patient's Node subsystems

Medical sensors Processor Memory Communication Power

2.3. The gateway node (GW) in this system is a coordinator in the WSN. Its functions are start the network, nodes management, and collecting the data from the WSN. 2.4. The Base station is processing the data that is received from the gateway and it display in the GUI. If the processing data are out of the range, the alarm will automatically send to the doctor node. In relation to the software used for processing system is LabVIEW. 2.5. WSN used to connect the whole nodes in this network which consists of GW and the nodes exists in the system (patient's nodes and doctor node).

3 SYSTEM IMPLEMENTATION
As shown in figure (1), we have chosen a combination of open source platforms as will be explained in the following paragraphs. 3.1. Patient Node. Refer to figure (2); this node consists of the following: MCU: ATmega328P ECG sensor. microcontroller. Heart rate (HR). Xbee communication. Temperature sensor Memory, 9V battery. (Temp). The hardware details of each of the mentioned component are as follows. ECG circuit ECG in Fig (4) Patient node components. XBee Thermistor in Thermistor

Doctor's Node subsystems

Alarm + LCD Processor Communication Power

Rest of the world

WSN

Fig. 1. System block diagram In the next paragraphs we will discuss each part of the block diagram. 2.1. The Patient nodes (sensing nodes) are consisting from subsystems: medical sensors, processor, memory, power, and communication. Each subsystem is presented in the following state: Medical Sensors subsystem: In healthcare, the sensors are medical sensors that combine transducers for detecting electrical, thermal, optical, chemical, genetic, and other signals to indicate the patient heath status *7+. Processor subsystem: it used to processing the data, scheduling the node task, and control the other subsystems. Memory subsystem: it used to store the data as file. Communication subsystem: it used to connect the nodes. Power subsystem: it supplies power to the node using a battery. 2.2. The Doctor node has the same patient node subsystem except there are no sensors. This node alarms the doctor in emergency case. So it contains an LCD and alarm sound.

ECG Electrodes

MCU

Fig. 2. Patient node components. 3.1.1. Sensors: ECG (Electrocardiogram): the ECG shows the electrical signal of the heartbeat which describes the heart health. There are many type of ECG, we used (leads II), the left side and the right side of chest, with one that is used as reference. The main leads are the left and right which measure the cardiac activity. These leads connecting to the patient body using electrodes. The electrodes (sensor) is a conductor which pass the electrical signal from the human body to the wires.The signal of the heart is low power between 1-5 mv and it contains a noise. So it must be processed using amplifiers and filters to be ready for the next stage (A/D conversion). WE can understand the (ECG) circuit from the block diagram in figure (1). The input of the circuit is the three electrodes that is low power noise signal, after the condition circuit the output signal will be (0-5 V) without noise.

15

The ECG sensor used in this sytem is olimex whose detail stages as shown in figure (3) *8+.

5V DC

2K ohm

Measured voltage

NTC

Sensors

High Voltage protection

HF rejection

Instrumental Amplifier

Filters

ECG

Fig. 3. ECG circuit stages. Fig. 5. Thermistor circuit. Heart Rate (HR): The number of heart beats per minute of time. The normal range of HR for adult is 60 -100 and for children from 60-140. This rate can be measured form the ECG; we have built an algorithm in the patient nodes to calculate the HR as shown in figure (4).
Start Threshold= const i=0

Read ECG

Yes
i = i +1

ECG > Thr

No

End of data

No

Yes
Heart Rate =i End

3.1.2. Processor: The processor used in this research is the Adruino platform as shown in figure (2) which has a MCU ATmega328P. The MCU features 14 digital input/output pins (of which 6 can be used as PWM outputs), 6 analog inputs, and 16 MHz ceramic resonator *10+. The Adruino is open-source platform based on flexible, easy-to-use hardware and software. The MCU controls the node sensing operation, after receiving the data from the GW to start this operation. It is starting the ECG circuit and temperature circuit. It is sampling the analog signal form ECG and thermistor and it is converted to 10-bit digital signal using Its A/D converter. The sampling rate for ECG is 200samples/sec. It's sending the sensing data from ECG, thermistor, and the HR to the GW for monitoring in GUI and alarm system. Setup operation is controlled by the microcontroller. In this operation the MCU receiving the patient personal data and store it in the node. Also the MCU performs the searching operation by sending the node address to the GW for making the list of in network nodes. 3.1.3. Communication: It used to connect the nodes in the network, we select XBee module in this research. This module has Transmit Power Output 2mW, RF Data Rate 250,000 bps, DSSS (Direct Sequence Spread Spectrum), and Operating Frequency Band ISM 2.4 GHz. It's Supported Network Topologies Point-to-point, Point-to-multipoint, Peer-topeer, and Mesh. We configure this XBee as end -device framework. *11+ 3.1.4. Memory Each patient node has SD card, for store the data in a log. The size of the SD card is 2GB, and the log type used is text file. The text file which store the data contained the personal data, temp, ECG, HR, date, and time. Every time the node sends the data to the GW, it stored in the SD card. So this card store data for long time and for multi patient. This text file can be displayed at any time using the computer. 3.2. Doctor Node. The doctor node consists of: MCU: ATmega328P LCD microcontroller. Alarm sound Xbee communication. 9V battery.

Fig. 4. Algorithm for the heart rate calculation. Temperature: the human temperature can be measured using biomedical sensors called thermistor. Thermistor properties that should be considered are tolerance, response time, and range. The normal human body temperature is between 35-37 degrees Celsius so you must choose sensor works in this range with good tolerance. We chose in this research a temperature sensors model (NTC- MA100BF103A). Its tolerance 0.05 in the temperature ranges 35-39 oC and its time response 2 second *9+. The signal condition circuit for thermistor is voltage divide with chosen resistor, and using the reference voltage from the ECG circuit to increase the resolution of the thermistor. The circuit for measured temperature is shown in figure (5).

16

The doctor node has the same patient node components but of course without sensors and it has 16*2 char LCD and (Piezo) alarm sound as shown in figure (6). The MCU performs the alarm operation, after the node receiving the data from the GW. After receiving the alarm information (patient node address, HR status, HR as a number, temp status, and temp as a number), the MCU display this information in the LCD and it called the voice of alarm. Finally the doctor node sends the success of the receipt of data to the GW.

is LabVIEW. This software performs three main operations, referring to figure (8) each operation is summarized as follows.
Start

transmitting control data Node wakeup

Doctor received button

Sound LCD

Processes inside each node

Received Control data Node Setup Node search

Control data Node Sensing

Receiving Personal Data Setup

Sending Patient data and sensing data Store data in SD card

Sending Node Address and status

Fig. 6. Doctor Node 3.3. Gateway node (GW). The coordinator or GW in this research is an XBee and its USB explorer circuit. This XBee uses ZigBee coordinator framework which it has all coordinator features like staring network, router, store, and manage information about the network including routing tables and security keys *12+. This node is passing the data form WSN to computer and vice versa. So it works as a gateway between WSN and the computer. 3.4. WSN We used ZigBee protocol and star network topology for the wireless connection. It is worked in 2.4GHz band, it is wider band providing more channels *6+, and its physical layer (PHY) supports 16 channels. We designed a network frame which contains the nodes selection, control operation, and packets type. The network frame is shown in figure (7).
Destination address Operation control

End of transmission Node sleep Display the Temp, ECG, and HR No Yes Send Data to Doctor

Alarm End

Fig. 8. Flowchart of software operation Setup operation: This operation for patient nodes only, to store the patient personal data at the MCU in the patient node, these data are Patient Name, Doctor Name, Age, Gender, Date, and Time. As shown in figure (9), it is easy way to setup any patient nodes.

Header

Type

data

stop

Fig.7. Network frame. Header: the protocol header. Destination address: node address. Type: packet type (data or control). Operation control: setup, searching, or sensing operations. Data: data bytes. Stop: End of packet 3.5. Base Station. The base station in this research can be a laptop or pc, because it uses USB for data communication and portable software. As previously described, the SW used

Fig. 9. Patient nodes setup GUI

17

Searching operation: It performs the searching for all nodes in the network, so the other operations can be processed on these nodes. The system can update these nodes at any time by starting the searching operation. Sensing operation: The setup and searching operations are supported the major operation which is the sensing operation. In the sensing operation, the base station sends a control packet to the patient nodes to start sensing, each node process the temp, ECG, and HR data and sends this data with the personal data back to the base station. The base station receives these data and processed it, and then displays each data (personal, temp, ECG, and HR) instantaneously in its display place in the GUI as shown in figure (10). In the same time, the base station scans the received data to give an alarm signal to the doctor for any emergency case. If the alarm system was ON, the doctor receives the data form the base station contained the node no, HR status (high or low according a threshold value depending on a medical standard), HR value, Temp status (high or low according a threshold value depending on a medical standard), and its value.

There are two modes in this operation; Multi Node or single Node. The software displays the message status for the operation or any error can be happened in the all system.

4 SYSTEM FEATURES
We developed this prototype system is characterized by the following features: Real time monitoring without delay. Multi patients. Auto detection for any new nodes. Portable alarm system with display. High resolution ECG monitoring. Auto detection for nodes failure and communication errors. Display system status for all operations Software can operate at any computer. Wireless system no more cable complexity. Free patient movement. Friendly user interface.

Fig. 10. The GUI the patient node. Personal data are stored in the node using setup operation before this operation. The ECG displayed is real-time ECG signal from patient node, also the HR and temperature. The system is tested under the supervision of the doctors in the hospital and after many program runs they tell us that the results displays on the GUI are very satisfactory with confidence.

5 RESULTS
The complete WSN layout in hospital is shown in figure (11). The designed healthcare system is shown in figure (12). All nodes connected with the Base station through the GW and the base station display data that received from the nodes. The system is installed and tested in Merjan Hospital Babylon city for many patients, the patient data that has been measured in hospital and the results is shown in figure (13). This figure shows the received data from

18

6 CONCLUSIONS AND FUTURE WORKS.

This work presents a prototype system for real-time healthcare monitoring. This system is designed due to the need for intensive care system that facilitates the work of medical staff. The system has very good feature such as a real time operation and contains the node doctor and other features wich increase the system confidence. This system has been tested in Merjan Hospital-Babylon city in the presence of specialists in the field of cardiology and Patient Node 1

intensive care and earned a certificate benefit of it. According to the obtained results and the certification of the doctors in charge in the hospital the presented healthcare system is very useful and applicable. As a future work, this system can be developed to be suitable to be applicable as home healthcare monitoring using internet, or it is possible to take advantage of this system in the house and make it an individual system using a single node and base station. These features may be our next research.

Doctor in remote area

WSN

GW

Base station

WSN

WSN

Patient Node 2

Base Station

Fig. 11. Intensive Care Unit layout in Merjan Hospital- Babylon.

19

+ Patient Node 1 Patient Node 2 Patient Node 3 Doctor Node

GW

Base station

WSN

Fig. 12. Prototype healthcare system.

Fig. 13. Results of a Patient in GUI *3+ Wan-Young Chung, Seong-Mo An and Seung-Chul Lee, "Real Time Multi-hop Routing Protocol for Healthcare System Based on WSN", South Korea, 2012. *4+ Bonam Kim, Youngjoon Kim, InSung Lee, Ilsun You, "Design and Implementation of a Ubiquitous ECG Monitoring System Using SIP and the Zigbee Network", Korea, 2007. *5+ Meng-Shiuan Pan, Yu-Chee Tseng," ZigBee Wireless Sensor Networks and Their Applications", Taiwan,

7 REFERENCES
*1+ Norsheila Fisal, Rozeha Abd Rashid, Mohd Adib Sarijari, Haslinah Mohd Nasir, " ECG Monitoring System Using Wireless Sensor Network (WSN) for Home Care Environment", Malaysia, 2009. *2+ Jin Soo Choi, "WIRELESS SENSOR NETWORK FOR HEALTH MONITORING", New Jersey Institute of Technology, 2012.

20

2008. Jukka Suhonen, Mikko Kohvakka, Ville Kaseva, Timo D. mlinen, Marko Hnnikinen, "LowPower Wireless Sensor Networks Protocols, Services and Applications", Springer, 2012. *7+ JeongGil Ko, Chenyang Lu, Mani B. Srivastava, John A. Stankovic, " Wireless Sensor Networks for Healthcare", IEEE, 2010. *6+

*8+ Olimex datasheet, https://www.olimex.com/dev *9+ "NTC THERMISTORS: TYPE MA", datasheet, http: //www.mouser.com/ds/0/GE-Sensing/ma-1122.pdf.

*10+ "ATmega48PA/88PA/168PA/328P", datasheet, http: //www.atmel.com/Images/doc8161.pdf. *11+ "XBee Series 2 OEM RF Modules", datasheet, ftp://ftp1.digi.com/support/documentation/90000866 _A.pdf. *12+ Yuan-Jen Chang, Chin-Hsing Chen, Li-Feng Lin, Ruo-Ping Han, Wen-Tzeng Huang, and Guey-Chuen Lee, "Wireless Sensor Networks for Vital Signs Monitoring: Application in a Nursing Home", Taiwan, 2012.