MOBILITY BASED SELF-ASSESSMENT MONITORING SYSTEM FOR DIABETES PATIENTS USING IOT

Transcription

1 Volume 118 No , ISSN: (on-line version) url: ijpam.eu MOBILITY BASED SELF-ASSESSMENT MONITORING SYSTEM FOR DIABETES PATIENTS USING IOT J.N.Maria Boncy 1, Dr. D.Shanthi 2, S.K.Somasundaram 3 1 PG Scholar, 2 Professor and Head, 3 Assistant Professor Department of CSE, PSNA College of Engineering and Technology, Dindigul. jn.mariaboncy@gmail.com, dshan71@gmail.com, soma@psnacet.edu.in Abstract Internet of Things (IoT) is a network of physical devices that has ability to transfer the data without human interaction. Self-assessment monitoring system is used to analyze the diabetic level of a patient themselves by collecting the data through sensors and storing it in the database. The main objective is to design and develop the mobility based self-assessment monitoring system for diabetes patients using IoT. The assessment is carried out by monitoring the patient s body sensor values (ECG, BP and blood glucose) in a web based link using IoT (Internet of Things). In 8051 microcontroller, the sensor values are collected via Analog to Digital Converter (ADC0808). These sensor values can be monitored only within range of 10 meter (30 feet) using Bluetooth device. In proposed system, arduino microcontroller kit is used to collect the sensor values via analog input pins. They have default ADC operation so there is no need of any third party devices like Analog to Digital Converter. In order to overcome these consequences of communication range, ESP8266 Wi-Fi module is used for communication and to monitor the sensor values widely. KNN classifier is used to classify the vital medical data and to enable the patients with high risk for regular treatment and to improve the quality of life of the patients. Key words: Arduino, Sensors, ESP8266 Wifi Module, Notification, Authentication and classification. I. INTRODUCTION Smart devices communicate with each other over the network is called Internet of things (IoT). The core of IoT is to achieve the exchange of information and communication between Things or smart objects including people, and perform a variety of information services and applications. Recently, a newer of IoT called Cognitive IoT has been enunciated. It aims at integrating cognitive technologies into IoT-based systems to ensure the smart management through enabling the cooperation and interaction between IoT and human. Coupling autonomic computing with cognitive computing sheds light on unprecedented opportunities for the development of smart IoT based systems. In 2008, about 347 million people in the world had diabetes. India had 69.2 million people living with diabetes as per the 2016 data. By 2030 there will be greatest increase of diabetes people in India which is been reported by diabetes association. Attention for diabetic patients is essential to maintain the blood glucose level. Through self-assessment monitoring system, the diabetic patients need not visit the doctors regularly. This Self-assessment monitoring system is used to analyze the diabetic level of a patient themselves by collecting the data through sensors and storing it in the database. Here patient s body sensors values are collected using (ECG, BP and Blood Glucose) and the values are transferred to database without any human or doctor interaction. The system is to design and develop the mobility based self-assessment monitoring system for diabetes patients using IoT. II. RELATED WORK Prosanta Gope et al [1] work propose that the low-power and lightweight wireless sensor nodes that are used to monitor the human body functions and surrounding environment. Each sensor node is integrated with biosensors such as Electrocardiogram (ECG), Electroencephalography (EEG), Blood Pressure (BP), etc. These sensors collect the physiological parameters and forward them to a coordinator called Local Processing Unit (LPU). In IoT-based healthcare application, the sensor nodes 3705

2 collect and forwards sensitive data to a coordinator. An adversary can eavesdrop on the communication, and can overhear critical information and also reduce computational overhead. Shih-Hao Chang et al [2] work proposes the system which offers convenience and lower the risk of erroneous measurement. In the basic rule service, the patient s standard blood-glucose range is divided into different diabetic types (such as Type 1 or Type 2) and measurement scenarios to produce seven grades of blood-glucose measurements (low 3, low 2, low 1, normal, high 1, high 2, and high 3). Because the entire blood-glucose monitoring measurement range is between 20 milligrams per deciliter and 600 mg/dl; the abnormality in either case can be critical. Because each patient s health condition is unique, the system allows patients to set their own grade ranges based on their doctor s suggestions, with the exception of low 3 and high 3. Blood-glucose grades will be applied to the anomaly detection rules to determine levels of abnormality. However this paper is impractical on large datasets and also unsuitable for actions with varying aspect ratios. Emna Mezghani et al [3] work proposes generic and reusable solutions for elaborating flexible smart IoT-based systems able to perceive the collected data and provide decisions. Based on the blackboard pattern, the Cognitive Monitoring Management pattern enables the interaction of IoT with human. It identifies a bidirectional interaction: IoT-Human interaction to visualize the data, extract new insights and receive notifications in case of context changes; and the Human-IoT interaction to manage the system through modifying its context and allow the IoT-based system learning from experts and acquiring knowledge. The drawback of this paper is heterogeneity of the knowledge representation and its distribution are identified as impediments of the integration. Taiyang Wu et al [4] work proposes a wearable sensor node with solar energy harvesting and Bluetooth low energy (BLE) transmission that enables the implementation of an autonomous WBAN. A web based smartphone application is also developed for displaying the sensor data and fall notification. To extend the lifetime of the wearable sensor node, a flexible solar energy harvester with an output based maximum power point tracking (MPPT) technique is used. The sensor node is integrated with an on board accelerometer, temperature sensor and a plug-in PPG sensor on a flexible solar panel. All the data from the sensor nodes and fall notification will be transmitted to a web-based smartphone application through a commercial BLE module. The drawback of this paper is high power consumption and transmits data only within short distance. Majid A. Al-Taee et al [5] work proposes a new BG pattern mining algorithm for more targeted therapeutic decision support in diabetes self-management. Based on patients BG readings collected via a handheld device and logged on a web-based health portal, the existing BG patterns are extracted in real-time and feed back to the patient along with appropriate therapeutic recommendations, educational modules and health care advice. Modules are linked by an existing telecommunications infrastructure based on GSM networks (3G/LTE) and/or Wi-Fi networks linked to a core IP network. The developed BG pattern mining algorithm has been deployed on an existing ehealth platform and linked to patients handheld devices. The drawback of this paper is data loss and limited distance communicated the patient data. Yin Zhang et al [6] work proposes digital data, including individual or clinical gene or protein data, can help identify the drug side effect and the new effect. This is the typical medical data, usually collected by medical service providers for clinical diagnosis such as EMR, and medical image. These data can be unified, managed, and opened to researchers with a necessary precondition for ensuring the privacy of the patient, to maximize the value of clinical medical data mining. An emotion-aware healthcare service promotes the innovation of modern medical with humanistic treatment. The drawback of this paper is lack of rapid response to emergency situations. Majid A. Al-Taee et al [7] work proposes self-management of diabetes enables real-time clinical interaction and feedback tailored to the personal needs of the patient, utilizing current and historical patient data. The physical layer nodes are linked to a Web based application layer through an existing telecommunication infrastructure. It interfaces the various objects of the physical layer to other objects. The application modules, which handle all users related functionality, are designed to be compliant with the Model-View-Control (MVC) pattern. Based on the collected reading, the mobile phone provides the necessary feedback and support in calculating the required insulin bolus. The drawback of this paper is lower efficiency and low manageable capability to cover over distance. Kasim M. Al-Aubidy et al [8] work proposes real-time monitoring and alarming system for patients. The system has an embedded microcontroller connected to a set of medical sensors and a wireless communication module. Each patient is considered as a node in a wireless sensor network and 3706

3 connected to a central node installed at the medical center through an internet connection. Microcontroller is interfaced with sensors and Bluetooth module for wireless communication with smart phone of the patient. Advanced wireless security techniques such as SSH secure connections is used to protect database and proxy to access information. The drawback of this paper includes loss in network interrupts communication. Udit Satija et al [9] work proposes a system to design and development of a light-weight ECG signal quality assessment method for automatically classifying the acquired ECG signal in real time based on IoT. A novelty based quality-aware ECG telemetry system for IoT-enabled cardiac health monitoring in wearable medical body area networks. The main objective is to present a light-weight real-time signal quality assessment technique for improving battery lifetime of IoT-enabled wearable devices and reducing the cloud server traffic load, bandwidth and treatment costs. The drawback of this paper is it cannot be spilt in a way that decreases connectivity cost. Istepanian.R.S.H et al [10] work proposes the potential benefits of using m-iot in non-invasive glucose level sensing and the potential m-iot based architecture for diabetes management. This technology enables new communication connectivity routes between mobile patients and care services through innovative IP based networking architectures. An emerging integrated technique will be explored and investigated towards generating a better performance in the physiological signal capture of the non-invasive glucose monitoring sensor This senor could also combine the accelerometer motion sensor within the integrated sensor cavity. It achieves a stand level of clinical diabetes assessment, and easy-to-operate and routine management. The drawback of this paper involves signal capturing, processing and transferring of sensor values. III. SYSTEM DESIGN The wearable sensor node terminals can be connected with other plug-in sensors such as ECG, Blood Pressure Sensor and Blood Glucose Sensor for other human body vital signals measurement in the proposed implementation. It is clear that the proposed system is accurate in sensing, clear in monitoring, intelligent in decision making and reliable in communication. Figure.1 represents the system architecture diagram. Fig.1.System Architecture Diagram 3707

4 In proposed system, we use the arduino microcontroller. The arduino microcontroller is an open source and user-friendly device. The proposed system consists of arduino, Blood Presser Sensor, ECG sensor, Glucose sensor, 16*2 LCD and ESP8266 wifi Module (IoT).We monitor the patients health by using Blood Presser Sensor, ECG sensor and Glucose sensor. The sensor values are given to the microcontroller. The ADC conversion takes place before it is been displayed on the LCD. The microcontroller displays the sensor values in the LCD which can be also be monitored in the webpage widely using IoT. ESP8266 Wi-Fi module device is used for a communication purpose. So, they can communicate the monitored sensor value worldwide. All the healthcare monitored values can be stored in corresponding web link and the patients are given notification in the form of diet, insulin intake physical activity and illnesses. Several features present in the software of Arduino which makes coding so easy and fast that is not possible with simple microcontroller. The dynamic changes in communication range are given by Wi-Fi module ESP8266 through which communication is done widely. It provides better communication and self-management of diseases over diabetes patients. IV. SYSTEM IMPLEMENTATION Arduino microcontroller and the sensors (ECG, BP and BS) act as transmitter. Where various test cases can be can be viewed in serial monitor that is COM port. The receiver end contains ESP8266 Wifi module and LCD display. The obtained test cases can be visualized graphically in ThingSpeak IoT platform. A. Arduino Uno Microcontroller The open-source Arduino IDE is used to write code on the sketch and the sketch is been uploaded on the arduino board. Arduino is used to collect the information from the 14 digital input/output pins and Analog PORT pins. It compiles the code before it is been uploaded. The arduino boards are featured with the serial communication or COM port for displaying the output of the sensors. After the compilation and uploading is verified. The output can be viewed through the serial monitor. B. Analog to Digital Conversion and Sensor Interface Arduino has six different Analog to digital conversion pin. These pins are connected to six different Analog output product sensors. They are named as (A0, A1, A2, A3, A4, and A5) Output from the sensor is Analog values which vary from 0 to These 1024 values can be converted into binary format. The arduino controller only supports the machine level language. The binary value of 0 to 1023 can be considered as an output voltage varies from 0 to 5V. In our process, we are using the three different sensors (ECG sensor, Blood Pressure sensor and Glucose blood sugar sensor. All the Sensors have a three different Pin outs. They are named as a VCC, GND, VOUT.VCC and GND is a supply power of the sensor. Where, VCC connected to the 5V of arduino, GND connected to the ground. VOUT provides the output of the sensor values of the test cases. The ADC on the Arduino is a 10-bit ADC meaning it has the ability to detect 1,024 (2^10) discrete analog levels. C. ECG Sensor Interface An electrocardiogram (ECG) sensor consists of three nodes namely positive, negative and neutral. All the values obtained from the sensors are converted to analog to digital. Range with VCC of 3.3v and bandwidth of Hz. D. Blood Pressure Interface Blood pressure sensor consists of positive, negative and ground connection. The even power supply 5V is been supplied through the dotted board. The input pin is connected to the arduino board to upload with the program and the output pin provides the output of the sensor values. 3708

5 E. LCD 16*2 display The output of the body sensor values are display in the LCD. All the output pins from the arduino are connected to the LCD to display the output values. When it comes to displaying of output on the LCD, the values are generated by the program which is been mentioned on the arduino IDE. F. IoT Monitoring and Connection Details The IoT monitoring is run based on the communication protocol operation. Arduino collect all the sensor value via analog port. They can be converted to digital signal and then they can be stored in arduino controller. This stored sensor values send to the ESP8266 Wi-Fi module through the operation of communication protocol. This way of connection is called TTL connection (Transistor- Transistor- Logic).In Arduino microcontroller, TX and RX (Pin number 0 and 1) are used as a communication port. These pins are used to send and receive the sensing datas to other devices. To monitor the sensor value, TX pin of the arduino controller is connected to the RX of the monitoring protocol device. To control any object or load device, TX pin of the controlling protocol device connected to the RX pin of the arduino. G. Classification The datas obtained from the arduino are converted into text file which is taken as input to the classifier. The datas are classified and the output is obtained in the target feature. Label produces normal and abnormality of diabetes patients. Based on this target feature message box is generated. In pattern recognition, the k-nearest neighbor algorithm is used for classification and regression. The input consists of the k closest training examples in the feature space in classification and regression. The output depends on whether k-nn is used for classification or regression.. k-nn classification, the output is a class membership. An object is classified by a majority vote of its neighbors, with the object being assigned to the class most common among its k nearest neighbors. V. EXPERIMENTS AND RESULTS Thus the system is fabricated and verified under various test cases. The output of various test cases is mentioned as follows. The Figure.2 represents the kit designed for mobility based self-assessment monitoring system for diabetes patients using IoT. Fig.2.Monitoring Kit The program in the arduino IDE is uploaded in the arduino board. Figure.3 represents the Arduino IDE where verification and uploading of the code is carried out in sketch. 3709

6 Fig.3.Arduino IDE The COM port or the serial monitor displays all the test case values of various patients. Figure.4 represents the arduino output or the COM port screen. The sending and response of the ESP8266 Instruction and the sensor values are displayed along with the test cases. Fig.4.Arduino output screen The test case values transmitted from the ESP8266 Wifi module are interfaced with the ThingSpeak IoT platform. A private platform is created to view the BP value along with date and time. Figure.5 represents the Blood pressure sensing value on the web link. Fig.5.Blood Pressure sensor monitoring using web link 3710

7 This channel with BS value of the patient along with sensor values is displayed. The test case values of Blood sugar sensor are monitored on the web link. The high risk patients can monitor their blood sugar level frequently and on regular basis. Figure.6 represents the blood sugar sensor on the web ink. Fig.6.Blood Sugar sensor monitoring using web link This channel field holds ECG test case value of the patients along with date and time which helps for instant analyzing and visualization. Figure.7 represents ECG sensor monitoring values using web link. The horizontal and vertical field represents the ECG value range and date. Through which regular health status can be monitored. Fig.7.ECG sensor monitoring using web link The data are classified and the output is obtained in the target feature. Label produces the normal and abnormality of the diabetes patients. Based on the target feature message box is generated. An object is classified by a majority vote of its neighbors, with the object being assigned to the class most common among its k nearest neighbors. Once the test feature is entered the data are matched along with train feature and the target feature is classified and produced as result. Figure.8 represents the classification in mat lab using KNN classifier. 3711

8 Fig.8.Classification Accuracy, sensitivity and specificity are evaluated for performance estimation. Figure.9 represents the performance estimation graph through which the accuracy, sensitivity and specificity among all three fields are evaluated based on test, train and the target features. The output is the property value for the test case features. This value is the average of the values of its k nearest neighbors. Fig.9.Performance Estimation VI. CONCLUSION In this system, mobility based self-assessment monitoring for diabetes patients using IOT is proposed. We have used various sensors to know the health status of the diabetes patient. Healthcare of the human body can be integrated with the sensor node or terminal, like ECG, BP, and BS. We have designed a system which can monitor the patients health in real time and gives us awareness about their health. It also monitors all the sensor values in web link through API keys produced by ThingSpeak. The existing work has lot of limitations like amplifying sound and limited communication range. Thus the proposed work uses arduino to reduce noise amplification and ESP8266 Wi-Fi module overcomes this limited range of communication and the user can monitor the sensor value in worldwide as well as monitor the sensor values in LCD display. 3712

9 REFERENCES 1. Prosanta Gope; Tzonelih Hwang, A secure IOT based modern healthcare system using body sensor network, IEEE Journals & Magazines,vol. 16,pp , Shih-Hao Chang; Rui-Dong Chiang; Shih-Jung Wu; Wei-Ting Chang, A context aware interactive M-health system for diabetics, IEEE Journals & Magazines,vol.18,pp , Emna Mezghani, Ernesto Exposito, and Khalil Drira, A Model-Driven Methodology for the Design of Autonomic and Cognitive IoT-Based Systems: Application to Healthcare, IEEE Transactions on Emerging Topics in Computational Intelligence,vol. 1,pp , Taiyang Wu; Fan Wu; Jean-Michel Redouté; Mehmet Rasit Yuce, An autonomous wireless body area network implementation towards IOT connected healthcare applications, IEEE Journals & Magazines, vol. 5, pp , Majid A. Al-Taee; Suhail N. Abood; Waleed Al-Nuaimy; Ahmad M. Al-Taee, Blood glucose pattern mining algorithm for decision support in diabetes management, 14th UK Workshop on Computational Intelligence (UKCI), IEEE Conference Publications,pp.1-7, Yin Zhang; Meikang Qiu; Chun-Wei Tsai; Mohammad Mehedi Hassan; Atif Alamri, Health-cps: healthcare cyper-physical system assisted by cloud and big data, IEEE Journals & Magazines,vol. 11,pp , Majid A.Al-Taee;Waleed Al-Nuaimy;Ali al-ataby;zahra J.Muhsin; SuhailN.Abood, Mobile health platform for diabetes management based on the Internet-of-Things, IEEE Jordan Conference on Applied Electrical Engineering and Computing Technologies (AEECT),IEEE Conference Publications,pp. 1-5, Kasim M. Al-Aubidy; Ahmad M. Derbas; Abdullah W. Al-Mutairi, Real-time patient health monitoring and alarming using wireless sensor network, IEEE Conference Publications, pp , Udit Satija; Barathram Ramkumar; M. Sabarimalai Manikandan, Real time signal quality aware ECG telemetry system for IOT based healthcare monitoring, IEEE Journals & Magazines,vol. 4,pp , Istepanian R.S.H. Senior Member, IEEE, S. Hu, Senior Member, IEEE, N. Y. Philip, Member, IEEE and A. Sungoor, The Potential of Internet of m-health Things m-iot for Non-Invasive Glucose level Sensing, Annual International Conference of the IEEE, Jianqing Wang, Member, IEEE, Takuya Fujiwara, Taku Kato, and Daisuke Anzai, Member, IEEE, Wearable ECG Based on Impulse Radio Type Human Body Communication, IEEE Engineering in Medicine and Biology Society,vol. 63,pp , Min chen; yixue hao; kai hwang; lu wang; lin wang, diseases prediction by machine learning over big data from healthcare communities, IEEE journals & magazines, vol. 5, pp , Naruto arai, Dairoku muramatsu, and Ken sasaki, Transmission Model of Human Body Communication Incorporating Size and Distance between the Two Electrodes of a Transmitter, Electronics Packaging (ICEP), International Conference, Amit sheth ; Utkarsham Jammu; Krishnaprasad Thirumaravan ; Tanvi Banerjee, Augmented personalized health: How smart data with IoTs and AI is about to change healthcare, Research and Technologies for Society and Industry (RTSI), 2017 IEEE 3rd International Forum on, Florian Kammuller and Juan C. Augusto and Simon Jones, Security and Privacy Requirements Engineering for Human Centric IoT Systems using efriend, Software Engineering Research, Management and Applications (SERA),IEEE International Conference,

10 3714