DIIPSDDMT: Design and Implementation of an Intelligent and Prognostic System for Diabetes using Data Mining Techniques
Diabetes is a chronic disease which occurs when the insulin production of pancreas is less than the body requirement or when the sufficient insulin is being produced by the pancreas but is not being used by the body effectively. Insulin regulates body sugar levels and keeps it normal. Diabetes is on
2025-06-28 16:32:09 - Adil Khan
DIIPSDDMT: Design and Implementation of an Intelligent and Prognostic System for Diabetes using Data Mining Techniques
Project Area of Specialization Artificial IntelligenceProject SummaryDiabetes is a chronic disease which occurs when the insulin production of pancreas is less than the body requirement or when the sufficient insulin is being produced by the pancreas but is not being used by the body effectively. Insulin regulates body sugar levels and keeps it normal. Diabetes is one of the fastest growing diseases in the world with its numbers rising from 108 million to 422 million in 1980 and 2014 respectively according to WHO. Diabetes comes with a package of other deadly diseases like increased heart attack risks, making body more prone to limb infections leading to amputations, kidney failure and diabetic retinopathy causing blindness attributing to 2.6 % blindness in the world. Diabetes has attributed to 2.2 million deaths alone in 2016 due to high blood glucose. This death rate because of diabetes calls for an intelligent diabetes prediction system development to reduce the death rate. Diabetes has three types which are type 1, type 2 and gestational diabetes. So, an intelligent web application system for the diagnosis of diabetes has been developed. It not only predicts if the patient is suffering from diabetes or not but also determine the type of diabetes the patient is suffering from, specifically if its type 1 or type 2 diabetes, as well as the calculates the risk of heart attack for the diabetic person in the next ten years according to Framingham risk score formula. Moreover, it will show the endocrinologists in the city of the patient with their clinic address and numbers so the patient can easily locate the doctor for further checkups, it will also provide the meal plans for diabetics with different calorie needs. The web application will also educate the users about the diabetes and its types and help them get better understanding of this disease, as well as it will guide people about the precautionary measures they can and should take to keep themselves safe and their blood glucose levels normal. The major advantage of this web application is the diagnosis of diabetes and the determination of type of diabetes and the calculation of heart attack risk so the user can take early precautions to avoid or reduce the risk of having a heart attack. It will help the user to take precautions to stay safe from the consequences of diabetes and provide the guidelines to maintain a healthy life style to increase the life span and will not have to perform extensive research to reach a consultant in the city of his residence. The patient will be able to easily access the system at any time which will allow him to enter his symptoms to predict diagnosis, check values to determine whether the patient is suffering from type 1 or type 2 diabetes and enter his stats to calculate heart attack risk. The web application will maintain his records and let him view his report, meal plans, precautionary measures and consultant details.
Project ObjectivesThe objectives of the web application are as follows:
- The diagnosis prediction of diabetes will be done, the class label is determined using the datamining algorithm.
- For generating missing values of glucose and insulin, this is done using machine learning algorithm.
- To determine the type of diabetes patient is suffering from, which will be done on the basis of the symptoms checked by the patient.
- For the calculation of heart attack risk, which is done using Framingham’s risk score formula.
- The users will be able to view the endocrinologists in the city of their residence without facing the trouble of surfring the internet for hours to find the contact details of a good endocrinologist.
METHODOLOGYDiabetes Prediction System (DIIPSDDMT) is a web based application that has some distinguished and renowned goals which are achieved as follows:
- Prediction using Data Mining Algorithm
- Determining Type of Diabetes
- Calculation of Heart Attack Risk
Prediction using Data Mining AlgorithmTo discover new information from already existing enormous amount of data, machine learning statistics under the umbrella of data mining is used for this purpose. Machine learning algorithm used for prediction in our research is Artificial Neural Network (ANN).
Artificial Neural Network
ANN is influenced from the human neural network, consists of interconnected neurons and works in the similar way like neurons do. ANN is used for the classification and prediction purposes, which was reason behind its preference in our research work.
Test and Train
Before using the algorithm for predicting class label, dataset was partitioned into train set and test set using split data operator in rapidminer. 70% of data was used for training and 30% for testing. After that, training set was used to model the process, predictions were made to test the model.
Determining Type of Diabetes
Predicted class label tells whether user is diabetic or not. For diabetic user, system further determines the type of diabetes he is suffering from. If class label is predicted to be positive, it is important to know the type of diabetes user is suffering from, take the necessary precautions and to get treated accordingly.
Diabetes type 1 and type 2 are the most common types, user might be suffering from. System displays a form with list of certain symptoms for further checking the type. Diabetic patient can also view the listed precautions and the diet plans that can be beneficial for his health.
Probability of Heart Attack
Heart attack being a life-threatening disease calls for its risk calculation after diabetes. Having diabetes, increases risk of developing heart diseases, infections, eye and skin problems, kidney disease, nerve pain and many others as it is a fatal disease. In comparison to others, heart problems require special attention as heart is the most important organ of human body. Due to this reason, our system also calculates the probability of having a heart attack through Farmingham risk score formula for the diabetic patients in the next ten years based on values of parameters entered by the user.
Benefits of the ProjectBenefits of our web application are:
- The main goal of our application is to facilitate the users by providing an accurate diagnosis of diabetes.
- It determines type of diabetes by using the symptoms.
- It calculates the risk of having heart attack in the next ten years for diabetic patients.
- It will also help the user to find the endocrinologist in his city of residence.
- It educates the user about diabetes and help sort his meal plans based on his calorie intake.
- It will also tell the patient about the necessary precautions he should take.
- In short, our application saves the cost and energy of users.
- It is efficient in terms of time as user gets to know the results within minutes instead of days.
- It is affordable and easily accessible.
- it is user-friendly and easy to use.
Hardware Requirements:
The system hardware specifications are:
- i5 core processor or above.
- At least 4 GB and above RAM for the client-side computers.
- 250MB and above space should be available on client-side computers.
Software Requirement:
The system software specifications are:
- Windows 7/8/10
- Language: PHP
- Database: MSSQL Server
- Tool: Rapid Miner Studio
Communications Requirement
The Communications requirements for the system are:
- NIC (Network Interface Card) – It is a computer hardware component that allows a computer to connect to a network. NICs may be used for both wired and wireless connections.
- Wifi device for high speed internet.
- TCP/IP protocol-Internet service provider to access and share information over the Internet.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 9000 | |||
| External Hard Disk | Equipment | 1 | 3000 | 3000 |
| Internet Device(WiFi) | Equipment | 1 | 6000 | 6000 |