Design of Self-Learning System for Diagnosing Health Parameters using ANFIS

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 03 | Mar -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 2401
Design of Self-Learning System for Diagnosing Health Parameters using
ANFIS
Anitha.A [1], Mahalakshmi.R [2], Vidhya.J3, Jayanthi.G4
Student[1,2,3]Assistant professor[4], Department of ECE, Parisutham Institute of Technology & Science, Tamil Nadu, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract –Nowadays, predictingofheartdiseasesiscritical
issue for humans. In previous, heart diseases cannot be
identified before it attacks. This paper describes the work of
self learning system for monitoring heart activities using
ANFIS. Here, heart diseases can be understood after reaching
abnormal state or in any fall detection stage. This paper
develops a method of classifying thedegreeofheartdiseasesin
patients using ANFIS. Neural networkwillpredictthetargeted
level and fuzzy-logic compares the targeted level with current
inputs and optimizes it. Here IoT plays a major role. The
predicted and optimized data about heart disease is posted in
the server for continuous information and it will be posted for
every 10seconds interval. Hence the preliminary stages of
heart diseases can be identified and recovered and also
current problems and future directions can be identified.
Key Words: ECG, Pulse Oximeter, Arduino UNO,
ThingSpeak server.
1. INTRODUCTION
Heart diseases mainly occur due toblockingofbloodvessels.
It includes coronary artery diseases, myocardial infarction,
bradycardia and tachycardia. This may be caused by high
blood pressure, obesity, smoking, poor diet. High blood
pressure results in 13% of CVD deaths, tobacco results in
9%, stoutness 5% and diabetes 6%. Coronary vascular
disease is avoidable by predicting it earlier. Human
computer interaction system is a study of relationship
between people and computers mediated information. The
future of human computer interaction system lies in how
intelligently these systems can take into account the user’s
context like Application areas, social organisation andwork.
Researchers on recognising the daily activities of peoplehas
progressed steadily but little focus has been devoted to
recognizing jointly activities as well as movements in a
specific activity.
1.1 EXISTING APPROACH
In the existing system, health care system was used to
monitor the physiological signal and current position of a
patient by automatic learning feature. Health care box was
used for automatic detection of ECG signals and position of
the patient. ECG acquisition module detects the signal and
transmits it to the health care centre. The position of the
patient position of patient can be recognized using an
outdoor precision GPS. Through health care centre, doctors
can assist the detailed health condition of a patient when the
patient reached the abnormal stage or fall, that notification
for rescuing them is send to healthcare box. Healthcare box
constantly record the status oftheECGsignal.Onceabnormal
stage is reached, it will give an emergency notification to
doctors and concerned family members. It has some
disadvantages like remotemonitoringofpatients,prediction
of diseases is not possible, heavy cost.
2. PROPOSED SYSTEM
In this paper, Neuro-Fuzzy based health diagnosis and
Arduino is used as a gateway to communicate to the various
sensors such as ECG sensor, Temperature sensor, and Pulse
Oximeter sensor. The microcontroller picks up the sensor
data and sends it to the network through WiFi and hence
provides real time monitoring of the health care parameters
for doctors. Initially, ANFIS algorithm is converted as
embedded C code and it is dumped in microcontroller.
Usually ANFIS has two parts: a) Training b) Testing. In the
training part, targeted level is set by giving some data as
input from the data set of LVM file. During testing, the
clinical input data from the patient body is compared with
the targeted values. The data canbeaccessedanytime bythe
doctor. In this, remote monitoring system is that the data as
to be securely transmitted to the destination end and
provision is made to allow only authorizedusertoaccessthe
data. User/doctor can access the data by logging in to the
html webpage. At the time of extremity situation in order to
prevent the patient from abnormal situation, a twitter alert
message is sent to the doctor through ThingSpeak server
(Internet of Things) which is connected to the controller. A
mobile applicationsoftware(ThingsView)isalsoprovidedto
check the health conditions of a patient continuously
through the mobile phone.

2.1 BLOCK DIAGRAM
Fig -1: Block diagram
2.2 BLOCK DIAGRAM DESCRIPTION
2.2.1 INPUTS
Pulse Oximeter
Pulse Oximeter indicates fastness of blood and
measures oxygen content in blood. It has IR
transmitter and a photo detector which indicates
the flow of oxygen as 0’s and 1’s.The ‘1’ will be
taken as 1023 because of 10 bit ADC usage.
ECG
In this, three electrode ECG is used.Oneelectrodeis
placed in left arm and another in right arm and the
other will be taken as a referenceelectrode. Theleft
arm and right arm electrodes will determine the
heart pulses.
Temperature sensor
It senses the temperature of patient’s body.
2.2.2 Instrumentation amplifier
The sensor values are given as an input to the IA. It
has two parts a) amplifier b) Filter. Filter will
remove unnecessary signals, amplifier will amplify
the filtered sensor values.
2.2.3 ADC
The amplified analog signal isconvertedintodigital
signal by using ADC. The outputwill be represented
as 0’s and 1’s which will be given to
microcontroller.
2.2.4 ESP8266
It is a wifi module. It has a firmer in which
operating system (OS) is installed. For checking
network connection of ESP8266, putty software is
used.
2.2.5 ThingSpeak server
It is free server and is used for posting the data and
also for monitoring the health values continuously.
For this, each person is havinga separateusername
and password to login.
3.METHDOLOGY
3.1 ANFIS
It refers to Adaptive NeuroFuzzyInferenceSystem.
Neural network has learning abilitylikesupervised
and unsupervised learning. This paper describes
backward propagation in supervised learning. In
backward propagation mainly uses multi-layer
perceptrons which includes input, hidden and
output layer. This algorithm uses a computed
output error to change the weight values in the
backward direction. The formula for sigmoid
activation is
1
f(x) = ---------
1 + e-input
4. Result
4.1 ThingSpeak Server output
Chart -1: Pulse Oximeter Chart- 2: ECG Sensor

4.2 Twitter Output
Fig 2: Twitter output
4.3 Snapshot
Fig 3: Implemented Circuit
5. CONCLUSION
Nowadays, heart disease plays a major role in human life
because of the lifestyle changes. This design of self learning
system for diagnosing health parameters using ANFIS will
predict the heart diseases before it severely attacks the
human. This device is successfully designed and verified.
REFERENCES
[1] G. Virone, A. Wood, may 2011, “Advanced wireless
network for health monitoring system”.
[2] “Innovative approach for wireless health monitoring
system using client-server architecture” Ms. Poonam
Agrawal ,Prof. S. P. Hingway,jun 2013
[3] Manisha Shelar,Jaykaran Sing, Jan 2013, “Wireless
Patient HealthMonitoring System”
Chart-3: Temperature Sensor

Design of Self-Learning System for Diagnosing Health Parameters using ANFIS

More Related Content

What's hot (20)

Similar to Design of Self-Learning System for Diagnosing Health Parameters using ANFIS (20)

More from IRJET Journal (20)

Recently uploaded (20)

Design of Self-Learning System for Diagnosing Health Parameters using ANFIS