VISVESVARAYA TECHNOLOGICAL UNIVERSITY

“JNANA SANGAMA”, BELAGAVI - 590 018.

2020 – 2021
A Project Report
(Phase - I)
on

“IoT Based Solutions For Health Care : Moving AI To The

Edge”
Submitted in partial fulfillment for the award of the degree of
BACHELOR OF ENGINEERING
in

ELECTRONICS & COMMUNICATIUON ENGINEERING

Submitted By

ARUN G M [1JB17EC012]
KARTHIK HADAGALI [1JB17EC033]
KISHORE KUMAR V [1JB17EC037]
NANDAN KUMAR M S [1JB18EC422]
Under the guidance of
Dr. Mahantesh K
Associate Professor, Dept. of ECE

DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING

SJB INSTITUTE OF TECHNOLOGY

#67,B G S HEALTH AND EDUCATION CITY
Kengeri, Bengaluru-560 060.
 ||Jai Sri Gurudev||
Sri Adichunchanagiri Shikshana Trust ®
SJB INSTITUTE OF TECHNOLOGY
BGS Health & Education City, Kengeri, Bengaluru-560 060.

DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING

CERTIFICATE
Certified that the project work phase – I entitled “IoT Based Solutions For Health Care :
Moving AI To The Edge” carried out by ARUN G M [1JB17EC012], KARTHIK
HADAGALI [1JB17EC033], KISHORE KUMAR V [1JB17EC037] and NANDAN
KUMAR M S [1JB18EC422] are bonafide students of SJB Institute of Technology in partial
fulfilment for the award of “BACHELOR OF ENGINEERING” in ELECTRONICS AND
COMMUNICATION ENGINEERING as prescribed by VISVESVARAYA
TECHNOLOGICAL UNIVERSITY, BELAGAVI during the academic year 2020 – 21. It is
certified that all corrections/suggestions indicated for internal assessment have been incorporated
in the report deposited in the departmental library. The project report has been approved as it
satisfies the academic requirements in respect of project work phase – I prescribed for the said
degree.

Dr. Mahantesh K Dr. K V Mahendra Prashanth Dr. Ajai Chandran C K

Associate Professor Professor & Head Principal
Dept. of ECE, SJBIT Dept. of ECE, SJBIT SJBIT
 ACKNOWLEDGEMENT

I would like to express our profound grateful to His Divine Soul Jagadguru Padmabhushan
Sri Sri Sri Dr. Balagangadharanatha Mahaswamiji and His Holiness Jagadguru Sri Sri Sri
Dr. Nirmalanandanatha Swamiji for providing me an opportunity to be a part of this esteemed
institution.

I would also like to express our profound thanks to Revered Sri Sri Dr. Prakashnath Swamiji,
Managing Director, SJB Institute of Technology, for his continuous support in providing
amenities to carry out this project (Phase – I) in this admired institution.

I express our gratitude to Dr. Ajai Chandran C K, Principal, SJB Institute of Technology, for
providing me excellent facilities and academic ambience; which helped me in satisfactory
completion of project work (Phase – I).

I extend our sincere thanks to Dr. K V Mahendra Prashanth, Professor & Head, Department of
ECE; for providing us invaluable support throughout the period of my project work (Phase – I).

I wish to express our heartfelt gratitude to my guide, Dr. Mahantesh K, for his valuable
guidance, suggestions and cheerful encouragement during the entire period of my project work
(Phase – I).

I express my truthful thanks to Dr. Mahantesh K & Mr. Bhaskar B, project coordinators, Dept.
of Electronics and Communication for their valuable support.

Finally, I take this opportunity to extend my earnest gratitude and respect to our parents, teaching
& technical staff of the department, the library staff and all my friends, who have directly or
indirectly supported me during the period of my project work (Phase – I).

Regards,
Arun G M
Karthik Hadagali
Ksihore Kumar V
Nandan Kumar M S
 DECLARATION

I hereby declare that the entire work embodied in this project report has been carried out under
the supervision of Dr. Mahantesh K in partial fulfilment for the award of “BACHELOR OF
ENGINEERING” in ELECTRONICS AND COMMUNICATION ENGINEERING as
prescribed by VISVESVARAYA TECHNOLOGICAL UNIVERSITY, BELAGAVI during the
academic year 2020 – 21.

Arun G M [1JB17EC012]
Karthik Hadagali [1JB17EC033]
Kishore Kumar V [1JB17EC037]
Nandan Kumar M S [1JB18EC422]
 Abstract

Adaptable Critical Patient Caring system is a key concern in developing countries like India.
Most of the local hospitals lack serving proper health service due to unavailability of
appropriate, easy and scalable smart systems. With an improvement in technology and
miniaturization of sensors, there have been attempts to utilize the new technology in various
areas to improve the quality of human life. One main area of research that has seen an adoption
of the technology is the health-care sector. The people in need of health-care services suffer
because of not getting the timely and appropriate solution and help for their problem and find it
very expensive, this is particularly true in developing countries.
As a result, this project is an attempt to solve a health-care problem currently society is
facing. The main objective of the project was to design a remote health-care system. It’s
comprised of three main parts. The first part being, detection of patient’s vitals using sensors,
second for sending data to cloud storage and the last part was analyzing using AI/ML and
providing the data for remote viewing. Remote viewing of the data enables a doctor or
guardian to monitor a patient’s health progress away from hospital premises.
The Internet of Things (IoT) concepts have been widely used to interconnect the
available medical resources and offer smart, reliable, and effective health-care service to the
patients. The aim of the project was to come up with a Remote Health Monitoring System that
can be made with locally available sensors with a view to making it affordable if it were to be
mass produced. Here we propose a generic architecture, associated terminology and a
classification model for observing critical patient’s health condition with machine learning and
cloud computing.
Hence the proposed architecture collects the sensor data through Raspberry Pi micro-
controller and relays it to the cloud where it is processed and analyzed for remote viewing.
Feedback actions based on the analyzed data can be sent back to the doctor or patient through
Email and/or SMS alerts in case of any emergencies. In addition, we have also analyzed distinct
IoT security and privacy features, including security requirements, threat models, and attack
taxonomies from the health care perspective. Further, we have used few hashing algorithm to
protect sensitive data from being exposed in case of any breach.
The aim of this project is to build an adequate system for health-care to serve critical
patients with a real-time feedback method. Combining with the ml models, this project may
serve as a smart health-care solution for the society.
v
 Table of Contents
Chapter Particulars Page
No
Acknowledgment iii
Declaration iv
Abstract v
Table of contents vi
List of figures vii
Chapter-1 Introduction 1-2
Chapter-2 Literature Survey 3-5
Chapter-3 Challenges,Objective and Scope 6-7
3.1 Challenges 6
3.2 Objectives 6
3.3 Scope 7
Chapter-4 Proposed Methodology 8-20
4.1 Raspberry Pi 8
4.2 Bio Sensors :- 9-11
1. Blood Pressure Sensor 9
2. Temperature Sensor 9
3. Electrocardiography 10
4. Pulse Oximeter 10
5. Glucose meter 11
4.3 LCD display 11
4.4 Website 12-13
4.5 Machine Learning in Health-care 14-20
 Introduction 14
 Applications 14
 Opportunities 14
 Machine Learning Model in our project 15
 Implementation of model 16
a.) Multiple Linear Regression Model 18
b.) Logistic Regression Model 19-20
Chapter-5 Conclusion 21
 References 22

List of figures
Figure Particulars Page
1. Introduction of IoT in health-care 12
2 Proposed Methodology 8
3 Raspberry Pi 8
4. Blood Pressure Sensor 9
5 Temperature Sensor 9
6 Electrocardiography 10
7 Pulse Oximeter 10
8 Glucose Meter 11
9 LCD display 11
10 Website Registration and login page 12
11 Health record Page 12
12 Consultation form Page 13
13 Past Health records Page 13
14 Machine Learning Model 15
15 Database for implementing model 16
16 Predicted disease using Multiple linear regression model 18
17 Predicted disease using Logistic Regression model 20
 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
CHAPTER 1
INTRODUCTION

Healthcare is given the extreme importance now a days by each country with the advent of the
novel corona virus. So in this aspect, an IoT based health monitoring system is the best solution for
such an epidemic. Internet of Things (IoT) is the new revolution of internet which is the growing
research area especially in the health care. With the increase in use of wearable sensors and the
smart phones, these remote health care monitoring has evolved in such a pace. IoT monitoring of
health helps in preventing the spread of disease as well as to get a proper diagnosis of the state of
health, even if the doctor is at far distance. In this paper, a portable physiological checking
framework is displayed, which can constantly screen the patient’s heartbeat, temperature and other
basic parameters of the room. We proposed a nonstop checking and control instrument to screen
the patient condition and store the patient information in web server. A remote health monitoring
system using IoT is proposed where the authorized personal can access these data stored using any
IoT platform and based on these values received, the diseases are diagnosed by the doctors from a
distance.

Department of Electronics and Communication Engineering. SJBIT Page 1

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
IoT
The Internet of Things(IoT) is the network of physical devices, vehicles, appliances and
gadgets embedded with electronics, software, sensors, actuators and connectivity which enables
these things to connect and exchange data, creating opportunities for more direct integration of the
physical world into computer-based systems, resulting in efficiency and reduced human exertions.

IoT in Health Care

IoT in a Healthcare is a heterogeneous computing, communication system of apps, websites
and devices that connects patient and health providers to diagnose, monitor, track and store vital
statistics and medical information.

Few examples of IoT in healthcare :

• Headsets that measures brain wave
• BP, Glucose and ECG monitors
• Sensors embedded in medical equipment
• Surgical robots and device implants
• Any wearable technology device

Department of Electronics and Communication Engineering. SJBIT Page 2

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
CHAPTER 2

LITERATURE SURVEY

1. IoT based Remote Patient Health Monitoring System :

In this paper, personal physiological data from the patient is collected that simulates fall
detection, heartbeat, temperature, humidity, toxic gas, air quality control, pressure. The readings are
collected in a simple cloud database and can be viewed remotely by a doctor or Healthcare giver. The
data can also be used in research on medical issues affecting the elderly or chronically ill. On the
security of the data, the database system is protected with Advanced Encryption Standard (AES). This
generates the secret key which can be used to decrypt the patients’ records ensuring that only
authorized personnel access the data. This safeguards the patients’ records from unauthorized users
and hackers who may want to intercept.

2. IoT based Smart HealthCare Kit :

The main idea proposed in this paper is to provide better and efficient health services to the
patients by implementing a networked information cloud so that the experts and doctors could make
use of this data and provide a fast and an efficient solution. The proposed model can also be deployed
as a mobile app so that the model becomes more mobile and easy to access anywhere across the globe.

3. The Internet of Things for Health Care: A Comprehensive Survey :

The discussion on several important issues such as standardization, network type, business
models, the quality of service, and health data protection is expected to facilitate the provide a basis
for further research on IoT-based healthcare services. This paper presents eHealth and IoT policies
and regulations for the benefit of various stakeholders interested in assessing IoT-based healthcare
technologies. In sum, the results of this survey are expected to be useful for researchers, engineers,
health professionals, and policymakers working in the area of the IoT and healthcare technologies.

4. Secure Telemedicine System :

In this paper, we listed security risks for telemedicine system and suggested the
countermeasures. Our proposed system STS is hybrid encryption with group key algorithm and

Department of Electronics and Communication Engineering. SJBIT Page 3

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
pseudo random function. It makes STS more efficient and secure. STS provides the information
sharing with group key algorithm and the designated encryption with the given secret values in
advance. In addition, STS provides mutual authentication and access control with verification
processes and the reversed hashed chain values. Data integrity is also guaranteed with the HMAC
verification.

5. Autonomous Systems for Telemedicine :

In this paper there presented solutions for autonomous telemedicine applications that can be
controlled through a Bluetooth connection to a computer or mobile phone. As well as shifting the
main locus of healthcare to the home, cloud computing continues to enable a step-change in overall
standards of public health, by enabling preventative interventions and promotion of healthier lifestyles
among society as a whole. This will benefit patients and payers by creating a population with less
illness and longer lives, and help healthcare providers focus their resources and skills more effectively.

6. Machine Learning based Health Prediction System using IBM Cloud as PaaS :
The main objective found in this paper is to provide better treatment using more advanced
technologies at very low cost. They started this project to bring out a good result in the hospitals to
serve the patient. They used some of the existed techniques and technologies to give a new shape in
the hospital and nursing sector. Most of the ml models accuracy varied from 80% to 92%. The lowest
accuracy obtained is 80%. An important finding in this project is the appropriate uses of machine
learning models for medical patients and categorical data manipulations.

7. Machine Learning Algorithms for Disease Prediction Using Iot Environment :

From this paper we learned how to provide real-time health monitoring and also the prediction
of disease over the internet. It can work base on real-time as well as synthetic training data. Prediction
accuracy of FRF algorithm is better than other learning approaches. The system also can provide the
alert when any criticalness 24*7. The proposed machine learning algorithm provides the best accuracy
than other machine learning approaches with minimum time complexity. Future studies consist of
implementation of such a system with parallel processing with high dimensional data using Hadoop or
cloud environment.

Department of Electronics and Communication Engineering. SJBIT Page 4

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``

8. Identifying Opportunities for AI applications in Healthcare - Renewing the National

Healthcare and Social Services :
This paper described the process of identifying opportunities for deploying artificial
intelligence to healthcare and social services nation-wide in Finland. The process included involving a
large number of stakeholders from a variety of backgrounds for idea generation and elaboration in
workshops using a design thinking method and resulted in 34 use cases and architectural designs
identifying AI capabilities needed in the components of these designs. In addition, the potential ones
of the use cases were selected for prototype development.Interestingly, the use of image analysis is
less visible in the listing of required AI capabilities

9. IoT Based Patient Monitoring System Using Raspberry Pi :

This project is developed as a prototype to monitor the health parameters of the patient i.e;
heartbeat and temperature using Raspberry Pi 3. They proposed continuous monitoring of the patient
conditions and store the patient's data in the server using the IoT concept. Heartbeat sensor and
temperature sensor are interfaced with Raspberry pi 3 via PIC microcontroller and LCD displays the
heartbeat and temperature of the patient continuously and data is stored in cloud using Wi-Fi.

10. Wearable Sensors in Healthcare and Sensor- Enhanced Health Information Systems: All
Our Tomorrows?
This paper summarizes current research in wearable sensors for healthcare and sensor-
enhanced health information systems. Intensive research efforts are undertaken in this field which
evolves rapidly, yet there still is a need for large-scale evaluation studies apart from several technical
and organizational obstacles to be overcome.

Department of Electronics and Communication Engineering. SJBIT Page 5

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
CHAPTER 3

CHALLENGES, OBJECTIVES AND SCOPE

3.1 CHALLENGES

 Standardization : To address device diversity, immediate efforts are required. A dedicated group
can standardize IoT-based healthcare technologies. standardization should consider a wide range
of topics such as communications layers and protocol stacks, including physical (PHY) and media
access control (MAC) layers, device interfaces, data aggregation interfaces, and gateway
interfaces.

 IoT Healthcare Platforms : Specialized platform, libraries and appropriate frameworks should be
built so that healthcare software developers and designers can make efficient use of given
documents, codes, classes, message templates, and other useful data.

 Web Application Development : Process of healthcare app development, the participation of an

authorized body or association of medical experts is typically required to ensure an app of
acceptable quality. In addition, regular updates on healthcare apps based on the due consideration
of recent advances in medical science are vital.

 The Quality of Service : Healthcare services are highly time sensitive and require QoS guarantees
in terms of important parameters such as reliability, maintainability, and the service level.

3.2 OBJECTIVES

 Better patient experience : From being connected to the health care system through IoT, doctors
can improve the diagnosis accuracy as they are getting all the necessary patient data at hand. In a
word, we can say that it allows monitoring patient continuously and remotely.

 Provide an accurate detection : By using this system, we can get accurate result based on patient
health. Moreover, it will be less error, collect data in less time and more accuracy than any human
performances.

 Shows the outcome of the treatment : By accessing patients health data in real time information
helps to make decision for the doctor on how the treatment is going on and what should do next.
Over all, this project will enable the physicians to utilize the results from data collection and
analyze that data in real time.

 Bridging the gap between doctor and patient : Health care is all about the patient so the need of
the patient always comes first but it is a matter of fact that most of the patient feel uncomfortable

Department of Electronics and Communication Engineering. SJBIT Page 6

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
to go to hospital or visit doctor’s chamber. In this way, this system creates a communication
between patient and doctor by providing the data.

3.3 SCOPE

 With the increasing use of cloud services combined with AI, IoT devices are getting smarter.

 These two technologies not only continuously records the information,but also analyse the data
using ML.

 Patients data encryption using enhanced security algorithms and protecting them against data
breaches.

Department of Electronics and Communication Engineering. SJBIT Page 7

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
CHAPTER 4
PROPOSED METHODOLOGY

Fig : Proposed methodology

Proposed model consists of
1. Raspberry Pi
2. Bio-Sensors
3. LCD Display
4. Website
1. Raspberry Pi : The Raspberry Pi is a low cost, credit-card sized computer that plugs into a
computer monitor or TV, and uses a standard keyboard and mouse. It also now supports hot plugging
using USB ports. It has 14 more GPIO pins. It has EEPROM readout support for the new HAT
expansion boards. Pictorial view of Raspberry Pi is given below. Raspbian is a Debian-based
computer operating system (OS) for Raspberry Pi . The code for further stages is written in Python in
the Nano text editor of Raspbian OS.

Fig : Raspberry Pi

Department of Electronics and Communication Engineering. SJBIT Page 8

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
2. Bio-Sensors : It is an analytical device, used for the detection of a chemical substance, that
combines a biological component with a physicochemical detector.
We are using 5 Bio-Sensors which are as follows,
i. Blood Pressure Sensor
ii. Temperature Sensor
iii. Electrocardiography
iv. Pulse Oximeter
v. Glucose Meter.

i) Blood Pressure Sensor : A device that measures the pressure of the circulating blood against
the arterial wall. The pressure sensor in electronic circuits is in the form of an integrated circuit
that acts as a transducer, that is, it replicates (in the form of an electrical signal) the signal it
receives as a function of imposed pressure.

Fig : Blood Pressure Sensor

ii) Temperature Sensor : Temperature sensor is a device which is designed specifically to

measure the temperature of body and its environment. LM35 is a precision IC temperature
sensor with its output proportional to the temperature (in °C). With LM35, the temperature can
be measured more accurately than with a thermistor. It also possesses low self-heating and does
not cause more than 0.1 °C temperature rise in still air. The operating temperature range is from -
55°C to 150°C.The LM35’s low output impedance, linear output, and precise inherent
calibration make interfacing to readout or control circuitry especially easy.

Fig : Temperature Sensor

Department of Electronics and Communication Engineering. SJBIT Page 9

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
iii) Electrocardiography : ECG records the electrical activity generated by heart muscle
depolarizations, which propagate in pulsating electrical waves towards the skin. Although the
electricity amount is in fact very small, it can be picked up reliably with ECG electrodes attached
to the skin. The full ECG setup comprises at least four electrodes which are placed on the chest
or at the four extremities according to standard nomenclature (RA = right arm; LA = left arm;
RL = right leg; LL = left leg). Of course, variations of this setup exist to allow more flexible and
less intrusive recordings, for example, by attaching the electrodes to the forearms and legs. ECG
electrodes are typically wet sensors, requiring the use of a conductive gel to increase
conductivity between skin and electrodes.

Fig : Electrocardiography(ECG)

iv) Pulse Oximeter: Sensor used to monitor the amount of oxygen carried in the body. It
measures the percentage of hemoglobin that carries oxygen. Heartbeat sensor provides a simple
way to study the function of the heart which can be measured based on the principle of psycho-
physiological signal used as a stimulus for the virtual reality system. The amount of the blood in
the finger changes with respect to time. The sensor shines a light lobe (a small very bright LED)
through the ear and measures the light that gets transmitted to the Light Dependent Resistor. The
amplified signal gets inverted and filtered, in the Circuit. In order to calculate the heart rate based
on the blood flow to the fingertip. MAX30100 is sensor used to measure Pulse rate.

Fig : Pulse Oximeter

Department of Electronics and Communication Engineering. SJBIT Page 10

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
v. Glucose Meter : Glucose sensors are used to measure glucose concentration in blood. Sensors
contain electrodes to measure the current generated by an enzymatic reaction usually between
glucose, an enzyme, and a mediator. Glucose oxidase is the enzyme used to measure glucose
construction in blood.Glucose reacts with the enzyme GOx(ox) and liberates current.This is
measured as the current by the electrode.

Fig : Glucose Meter

3. LCD Display : The Raspberry Pi Touch Display is an LCD display which connects to the
Raspberry Pi through the DSI connector. In some situations, it allows for the use of both the HDMI
and LCD displays at the same time (this requires software support).

Fig : LCD Display

Department of Electronics and Communication Engineering. SJBIT Page 11

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
4. Website : It is designed to provide patient past health records and get Diagnosed from
specialized doctor.

a. Website Registration and login

i. Here first time patients need to enter their details for registration.
ii. After successful registration, patient can sign-in using e-mail and password.

iii. After successful login, patient can fill consultation form to get advise from doctors and fetch
previous health records in this webpage.

Department of Electronics and Communication Engineering. SJBIT Page 12

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``

iv. Patient needs to enter the reason for consultation and health parameters in case IoT kit is
not available.

v. Patient can fetch their previous health records and analyze their health condition.

Department of Electronics and Communication Engineering. SJBIT Page 13

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
Machine Learning in Healthcare
ML and AI technologies have been recently penetrating all spheres of healthcare services,
from improving healthcare management to new drug discovery. Even though it is unlikely that
computers will completely replace doctors and nurses, modern technologies are already transforming
the healthcare industry as we knew it.
Only in 2016, over 100 companies were offering machine learning algorithms and predictive analytics
to improve the effectiveness of the drug discovery process, provide assistance to patients, and
diagnose ailments by processing medical images.

Machine Learning applications that have emerged in the recent years.

1. Disease Identification/Diagnosis
2. Personalized Treatment
3. Smart Electronic Health Records
4. Behavioral Modification
5. Drug Discovery/Manufacturing
6. Clinical Trial Research
7. Epidemic Outbreak Prediction

Opportunities for Machine Learning in Healthcare

Health Catalyst believes the introduction and widespread use of machine learning in
healthcare will be one of the most important, life-saving technologies ever introduced. We
believe the opportunities are virtually limitless for the technology to improve and accelerate
clinical, workflow, and financial outcomes.

Reduce re-admissions : Machine learning can reduce readmissions in a targeted, efficient,

and patient-centered manner. Clinicians can receive daily guidance as to which patients are
most likely to be readmitted and how they might be able to reduce that risk.

Prevent hospital acquired infections (HAIs) : Health systems can reduce HAIs, such as
central-line associated bloodstream infections (CLABSIs)—40 percent of CLABSI patients
die—by predicting which patients with a central line will develop a CLABSI. Clinicians can
monitor high-risk patients and intervene to reduce that risk by focusing on patient-specific
risk factors.

Reduce hospital Length-of-Stay (LOS) : Health systems can reduce LOS and improve other
outcomes like patient satisfaction by identifying patients that are likely to have an increased LOS and
then ensure that best practices are followed.

Predict chronic disease : Machine learning can help hospital systems identify patients with
undiagnosed or misdiagnosed chronic disease, predict the likelihood that patients will develop chronic
disease, and present patient-specific prevention interventions.
Department of Electronics and Communication Engineering. SJBIT Page 14
 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``

Reduce 1-year mortality : Health systems can reduce 1-year mortality rates by predicting the
likelihood of death within one year of discharge and then match patients with appropriate
interventions, care providers, and support.

Predict propensity-to-pay : Health systems can determine who needs reminders, who needs financial
assistance, and how the likelihood of payment changes over time and after particular events.

Predict no-shows : Health systems can create accurate predictive models to assess, with each
scheduled appointment, the risk of a no-show, ultimately improving patient care and the efficient use
of resources

Machine Learning methodology in our project

Fig : Machine Learning Model

In the proposed ML model, the database is partitioned into two different parts, training and testing.

Training
 Collect data from the online repository like synthetic data as well as real-time patient data.

 Apply data mining approaches like data pre-processing, data cleaning, data acquisition, outlier

detection, and data conversion.

 Once complete these phases, data is saved into the database called as background knowledge,

which is used at the time of testing.

Department of Electronics and Communication Engineering. SJBIT Page 15

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
Testing
 Initially, IoT based environment is created by using sensors as wearable devices.

 Then these sensors are connected to Raspberry Pi, where data will be collected using a batch

processing approach from these sensors.

 All collected data then stored into a global database using a connection-oriented architecture.

 In testing, we read all training and testing data simultaneously.

 To predict the possible result using a decision-making approach.

 At the final stage provide accuracy analysis using True positive (TP ) and false negative (TN)

statements of system.

Implementation of Machine Learning in our project

In this project, we have predicted whether a person has diabetes or not, using an available data
of 768 patients inspecting their Glucose, BP, Skin Thickness, Insulin, BMI, Diabetes Pedigree
Function and Age.

The data was collected and made available by "National Institute of Diabetes and Digestive
and Kidney Diseases" as part of the Pima Indians Diabetes Database. Several constraints were placed
on the selection of these instances from a larger database. In particular, all patients here belong to the
Pima Indian heritage (subgroup of Native Americans), and are females of ages 21 and above.

We have used Python as our programming language, and making use of some popular python
machine learning and data science related package we have handled missing term in the dataset for
further processing.

Department of Electronics and Communication Engineering. SJBIT Page 16

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``

We have used Multiple Linear Regression algorithm and Logistic Regression algorithm to
build our classification model and predict the outcome..

Department of Electronics and Communication Engineering. SJBIT Page 17

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
Multiple Linear Regression model : It is a Supervised Learning algorithm where the predicted
output is continuous and has a constant slope. It is used to predict values with in a continuous range.
In our case the range is between 0 and 1. from Fig (a) if the predicted value nearing to 1 or greater
than 0.5 then it is declared as the person is suffering diabetes.

Department of Electronics and Communication Engineering. SJBIT Page 18

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``

Fig (a). Comparison of actual value with the predicted value using Multiple Linear
Regression Model.

Logistic Regression Model : It is used to predict events like pass/fail win/lose 0/1. it is a binary
dependent variable. We can clear justify from comparing actual and predicted value from Fig(b) that
if the value is 1 then person is suffering from diabetes.

Department of Electronics and Communication Engineering. SJBIT Page 19

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``

Fig (b). Comparison of actual value with the predicted value using Logistic Regression
Model.

Department of Electronics and Communication Engineering. SJBIT Page 20

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
CHAPTER 5

CONCLUSION

 To provide health-care assistance to aged and physically challenged, who require regular checkups
and to people in remote areas where basic health-care amenities are lagging.
 Using our IoT kit and user friendly website they can check different health parameters.
 Patients can monitor their health records and act accordingly based on online doctors consultation.
 Further we have used ML to predict and diagnose the disease at an early stage or even before the
disease actually develops based on symptoms.

Applications

 To analyze patients past data and to compare it with present condition in order to find any
deviations using AI/ML.
 Common diseases like fever, cold, cough etc can be diagnosed remotely.
 Continuous monitoring of patients suffering from chronic diseases like diabetes and pulmonary
diseases.
 On-time alert to patients and their emergency contacts is critical in event of life threatening
circumstances.

Department of Electronics and Communication Engineering. SJBIT Page 21

 IoT Based Solutions For Health Care : Moving AI To The Edge 2020-21

``
REFERENCES

[1] IoT based Remote Patient Health Monitoring System, Narasimha Rao Jasti Madhu, 2017 - Kansas
State University Manhattan, Kansas

[2] IoT based Smart HealthCare Kit, Punit Gupta, Deepika Agrawal, Jasmeet Chhabra, Pulkit Kumar
Dhir, 2016 - ResearchGate

[3] The Internet of Things for Health Care: A Comprehensive Survey, S. M. Riazul Islam, Daehan
Kwak, MD. Humaun Kabir, Mahmud Hossain, Kyung Sup Kwak, 2015 - IEEE

[4] Secure Telemedicine System, Hyun-A Park, 2018 - International Conference on CSCI

[5] Autonomous Systems for Telemedicine, Cornel Popescu, 2015 - 9th International Conference on
Complex, Intelligent and Software Intensive Systems

[6] Machine Learning based Health Prediction System using IBM Cloud as PaaS, Asif Ahmed Neloy,
Sazid Alam, Rafia Alif Bindu and Nusrat Jahan Moni, 2019 - ICOEI

[7] Machine Learning Algorithms for Disease Prediction Using Iot Environment, Samir S. Yadav,
Shivajirao M. Jadhav, 2019 - IJEAT

[8] Identifying Opportunities for AI applications in Healthcare - Renewing the National Healthcare
and Social Services, Pasi Tyrväinen, Minna Silvennoinen, Karoliina Talvitie-Lamberg, Anniina Ala-
Kitula, Reija Kuoremäki, 2018 - IEEE

[9] IoT Based Patient Monitoring System Using Raspberry Pi, Mohd. Abdul Muqeet, Mohammed
Umair Quadri, 2019 - ResearchGate

[10] Wearable Sensors in Healthcare and Sensor-Enhanced Health Information Systems: All Our
Tomorrows?, Michael Marschollek, Matthias Gietzelt, Mareike Schulze, Martin Kohlmann, Bianying
Song, Klaus-Hendrik Wolf, ResearchGate - 2012

Department of Electronics and Communication Engineering. SJBIT Page 22