Diagnosis and Classification of Diabetes Mallitus using Bio-Signals

Project Title

Diagnosis and Classification of Diabetes Mallitus using Bio-Signals

Project Area of Specialization Biomedical EngineeringProject Summary

Globally a large number of people are suffering from diabetes. In 1990, 11.3 million people were
diabetic across the world which has increased up to 22.9 million people in 2017. Almost 46% of
people are unaware of that they are diabetic which in turn cause serious complication. It is on the
list of the first ten deadly diseases in most countries. Its rate is increasing alarmingly worldwide.
According to statistics, it would be almost 366 million people in 2030. There are three main
types of diabetes named as Type 1, Type 2, and Gestational diabetes. In the proposed method,
the aim is to design a classification system that can distinguish between diabetic and non-diabetic
patients accurately using a non-invasive technique. The proposed method will use PPG as signals
and using ML and DL classifiers such as SVM, KNN, etc. it will distinguish between healthy
and diabetic individuals.

Project Objectives

A biomedical signal-based project that detects diabetes in a person and can also tell us that what
type of diabetes is present in that person. So that according to type cure can be recommended
like if it is type 1 then use insulin if it types 2 then physical activities such as weight loss,
prevention, and healthy diet, etc. are required. If diabetes is diagnosed early then it can control
and the patient is not much affected. So, it is compulsory to have more and more work on it. Our
main objective is to analyze the data given and make a system that can analyze and detect that
subject as healthy or diabetic. Then if the subject is diabetic then what type it is. To measure and
analyze accurately and efficiently we are developing a system having software as well as
hardware.

Project Implementation Method

The first step of the project is to acquire data through Bio-ensors (ECG and PPG). The data is taken both from online sources and in real-time. The data is then cleaned and the region of interest from the bio-signals is extracted. Through which the features are obtained. The machine learning models are trained on the resultant feature file and we get the acuuracy of diagnosis. 

Benefits of the Project

•  To have such a system that  work accordingly to detect diabetes efficiently without any error or mistake.
• This system will perform accurately and efficiently regardless of any computational complexity.
• This system will be inexpensive and easily accessible.
• It is easy to use and shows 100% accurate results without the need for any external device or any technical person.
• The system will measure and
  analyze accurately and efficiently

Technical Details of Final Deliverable

Raspeberry Pi 3b+ is used as the operating and standalone system to diagnose the diabetes. The real time data is obtained using the two sensors called: MAX30100 for PPG and AD8232 for ECG. Arduino Uno is used as the Analog to Digital (ADC) converter for the analog sensors. The LCD will display all the details of data being obtained and trained on. The final results of the diagnosis of the diabetes will be displayed there.

Final Deliverable of the Project HW/SW integrated systemCore Industry HealthOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Good Health and Well-Being for PeopleRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	46750
Raspberry pi 3b+ Processor	Equipment	1	20000	20000
MAX 30100 PPG Sensor	Equipment	1	1000	1000
AD 8232 ECG Sensor	Equipment	3	2000	6000
Arduino Uno	Equipment	2	2000	4000
LCD Display	Equipment	1	10000	10000
Report Printing	Miscellaneous	3	250	750
Circuit Components	Equipment	1	5000	5000