MULTIPLE DISEASE PREDICTION AND MEDICAL

CHECK UP USING MACHINE LEARNING

NAME : ARUNPRASATH J
ROLL NO: RA2232241010131
CONTENTS

❏ ABSTRACT
❏ LITERATURE STUDY
❏ SYSTEM ANALYSIS
EXISTING SYSTEM

PROPOSED SYSTEM

FEASIBILITY STUDY

REQUIREMENT SPECIFICATION

LANGUAGE SPECIFICATION - PYTHON

❏ SYSTEM DESIGN
SYSTEM ARCHITECTURE

USE CASE DIAGRAM

ACTIVITY DIAGRAM

SEQUENCE DIAGRAM

CLASS DIAGRAM

❏ MODULE DESCRIPTION
❏ TESTING
❏ CONCLUSION
ABSTRACT

➢ The project aims to develop a Random Forest Based system for disease analysis using
relevant data like age and blood glucose levels.

➢ Consists of three main components

1.Data preparation
2.Model training
3. Predictions

➢ Data preparation involves:

Collecting, cleaning, preprocessing and separating idea

➢ Model training focuses:

Selecting appropriate models and evaluating their performance
➢ Prediction module utilizes the trained system to predict the presence or
absence of disease based on new input data, presenting results ina
comprehensible format.
LITERATURE STUDY

➢ Deep learning algorithms show promise in medical diagnosis, with

accuracy comparable to healthcare experts,

➢ A novel conventional neural network (CNN) - based multimodal illness

risk prediction accuracy of 94.8% and faster convergence speed than
the CNN- based unimodal algorithm.

➢ Machine learning methods like support vector machine (SVM), knearest

neighbor (KNN), naive bayes (NB) , and decision tree can be used for
diagnosis of major disorders like cancer,diabetes, epilepsy, and heart
attack.
➢ Deep neural networks are increasingly used in academia and
business for image analysis and natural language processing.

➢ Heart disease prognostication aids like the Enhanced Deep Learning

aided Convolutional Neural Network (EDCNN) incorporate
regularization learning strategies and run on the loMT platform.

➢ K-Nearest Neighbor and CNN algorithms are employed to accurately

forecast diseases, with CNN outperforming KNN in predicting
diseases by 8.5 percentage points.
 SYSTEM ANALYSIS

1.EXISTING SYSTEM
Disease prediction systems are crucial for people’s lives, but traditional approaches
are not highly accurate or efficient, potentially leading to false assumptions and improper
diagnosis and treatment.

2.PROPOSED SYSTEM
● The proposed system uses machine learning and deep learning algorithms to predict

cancer, diabetes and pneumonia.

● It uses a stepwise approach, including cleaning, feature extraction, and classification, to
identify relevant elements.

● CNNs, a class of deep neural networks, are used for image recognition, natural language
processing, drug discovery, and health risk assessments.

● The system uses inputs like radius mean perimeter area, blood glucose, chest pain, age
and gender for cancer, diabetes, hearts disease, and malaria.

● After training, a rule set is formed, and the system classifies and predicts the most likely
disease.
ADVANTAGES OF PROPOSED SYSTEM

➢ Accurately classifies diseases using sample test data, providing

accurate results in real time.

➢ It is beneficial for short term data processing and instant results.

➢ Medical professionals can understand age related health risks,

warm patients, and prevent heart disease.

➢ Patients can consult doctors and undergo checkups, building

trust and security among people.
AFTERMATH, CAUSES OF THE DISEASE

➢ A study conducted in The UK and The USA aimed to investigate the impact of
medical errors on clinician well being and patient care quality.

➢ The study involved a cross- sectional survey of 265 physicians and nurses.

➢ The results revealed that errors can lead to professional or personal disruption,
emotional responses, coping strategies, and perceptions of organizational
support.

➢ The study also proposed an end to end application to analyze and predict
diseases, including liver, cardiovascular,kidney-related cancer, pneumonia and
malaria.

➢ The application would function as a virtual assistant in clinical laboratories,

healthcare centres, and medical clinics.
FEASIBILITY STUDY

➢ A business proposal outlines the primary goals and preliminary cost estimates
for a project aimed at improving server performance.

➢ A comprehensive study is necessary to assess the system’s viability.

➢ The feasibility research includes mostly three lines of thought:

1.ECONOMICAL FEASIBILITY

2.TECHNICAL FEASIBILITY

3.OPERATIONAL FEASIBILITY

4.SOCIAL FEASIBILITY
ECONOMIC FEASIBILITY
➢ The study suggests that using the open-source and free technologies can lead to
cost savings for upper management, as resources are limited and every dollar spent
must have a valid reason.

TECHNICAL FEASIBILITY
➢ This research aims to determine the system’s technical feasibility, ensuring smooth
development without excessive IT staff pressure, and ensuring simplicity in design
for minimal adjustments.
OPERATIONAL FEASIBILITY
➢ Research emphasizes the importance of user input in technology adoption.

➢ Training and orientation are crucial for quick adoption and fostering trust, as users
need to feel comfortable submitting constructive feedback without fear of threats.

SOCIAL FEASIBILITY
➢ The social feasibility analysis assesses the project’s potential to impact the
community, considering cultural norms and institutional frameworks that may
limit the availability of certain workers.
REQUIREMENT SPECIFICATIONS

HARDWARE REQUIREMENTS
PROCESSOR : Pentium Dual Core 2.00HZ

HARD DISK : 120 GB

RAM : 2GB (minimum)

KEYBOARD : 110 keys enhanced

SOFTWARE REQUIREMENTS
OPERATING SYSTEM : Windows 7 (with service pack 1),8,8,8 and 10

LANGUAGE : Python
LANGUAGE SPECIFICATION - PYTHON
➢ Python is a popular programming language due to its user-friendliness,
rich features, and versatility.

➢ It is ideal for machine learning, a branch of AI that enables computers to

learn from mistakes and perform routine tasks automatically.

➢ Python’s popularity stems from its ability to function on its own platform
and its active python community,which facilitates discussions and
suggestions on improving program functionality.

➢ This development is crucial for intelligent solutions to real world

problems.
SYSTEM DESIGN

SYSTEM ARCHITECTURE
➢ The diagram shows how many actions andchoices are linked together

➢ You might say that the whole process and how it was carried out is a picture.
ARCHITECTURE DIAGRAM
USE-CASE DIAGRAM
➢ The possible interactions between the user, the dataset, and the algorithm are often
depicted in a use case diagram.

➢ Advantages:

1.Tools for modelling a business

2.gathering requirements

3.Identifying system class

4.Identifying test during analysis,design

5. Participants understand the business’s operations, workers, customers and activites

USE CASE DIAGRAM:
ACTIVITY DIAGRAM

➢ `An activity diagram, in its most basic form, is a visual representation of the
sequence in which tasks are performed.

➢ It depicts the sequence of operations of that makeup the overall procedure.

ACTIVITY DIAGRAM
SEQUENCE DIAGRAM

➢ These are another type of interaction based diagram used to display the
workings of the system.

➢ They record the conditions under which objects and processes cooperate.
SEQUENCE DIAGRAM
CLASS DIAGRAM

➢ In essence this is a “context diagram”,another name for a contextual diagram.

➢ A + indicates a publicly accessible characteristic or action

➢ A - a privately accessible one

➢ A # a protected one

➢ A - denotes private attributes or operations.

CLASS DIAGRAM
MODULE DESCRIPTION

MODULE 1: DATA PREPARATION

➢ This module discusses the importance of data preparation in data analysis and
machine learning projects, focusing on the various steps involved in cleaning,
transforming, and manipulating raw data for accurate and reliable results.
➢ CATEGORISATION:
Data collection
Data cleaning
Data preprocessing
Data integration
Data reduction
Data splitting
MODULE 2: MODEL TRAINING
➢ Once the input data has been prepared on Module 1, the next step is to train machine learning
models that can make predictions based on the input variables.
➢ Steps involved:

Model selection

Data splitting

Model fitting

Hyperparameter training
MODULE 3: PREDICITONS
➢ Once the random forest models have been trained in Module 2, the next step is
to use the models to make on new,unseen data.
➢ In this we use four trained models to predict the presence or absence of each of
the four health conditions
Input data
Model predictions
Output results
MODULE 4 : WEB APP DEPLOYMENT
➢ This describes a python script to handle server side processing for a web
application.
➢ The code will be developed using flask, a tool that simplifies web application
creation.
➢ A virtual environment is created, and a new file called “main.py” is created.
➢ The script is executed, and the results are displayed on the first local server
hosted webpage.
➢ Flask’s route decorator is used to to link a URL to a function, and a new feature
called “add url rule” is added to associate a URL with a specific function.
MULTI-PAGE CREATION
➢ Alzheimer’s, cancer, heart disease, renal disease, and pneumonia are only few of
the diseases that we stimulate in various variants.
➢ Next , using the flask module and python notebook, we compile everything into a
single HTML,file.
TESTING

➢ Discovering and fixing such problems is what testing is all about.

➢ The purpose of testing is to find and correct any problem with the final
product.
➢ TYPES OF TESTING:
UNIT TESTING
INTEGRATION TESTING
FUNCTIONAL TESTING
TEST CASES
1. Code Coverage Testing: Test that all code paths and functions in the system are
executed and tested, including the image preprocessing pipeline, the model
architecture, and the prediction algorithm.

2. Unit Testing: Test individual units or components of the system, such as the image
preprocessing algorithms or the logistic regression classifier.

3. Integration Testing: Test how different components or systems of the system work
together, such as the integration between the convolutional neural network and the
logistic regression classifier
CONCLUSION

➢ In conclusion, a web application for predicting diseases using machine learning

and deep learning can provide a valuable service to health care patients and
professionals, by helping them to be protective and proactive towards the
condition.
➢ It can be , by providing increased accuracy, time-saving, user-friendliness, and
free access. By improving the accuracy, real-time numerical and image inputs can
be obtained and diseases can be predicted in real time wIth efficent and improved
models.
OUTPUT