THE SECOND SYNOPSIS OF THE PROJECT ENTITLED

HEART DISEASE PREDICTION SYSTEM

Submitted to:
GURU NANAK DEV UNIVERSITY,
AMRITSAR

SUBMITTED IN PARTIAL FULFILLMENT OF THE

REQUIREMENTS FOR THE DEGREE OF
MCA(FYIC)6TH SEMESTER(2024)

Supervised By: Dr. Sarveshwar Bharti Submitted By: Kanika Kumari

Roll No. : 27502100065

Department of Computer Science

Guru Nanak Dev University, Amritsar
 Table of Content

S.No. Content Page No.

1. Introduction 1

2. Problem Statement 2-3

3. Requirement Analysis 4-5

4. System Specification 6-8

1. Introduction

Heart disease is a leading cause of mortality worldwide, with various factors contributing to its
prevalence. Early detection and prediction of heart disease play a crucial role in providing timely medical
interventions and improving patient outcomes. In this project, we aim to develop a heart disease
prediction system utilizing machine learning techniques.

The primary objective of this project is to train a predictive model capable of accurately assessing
the likelihood of an individual developing heart disease based on a set of relevant features such as age,
gender, blood pressure, cholesterol levels, and other medical indicators. To accomplish this, we employ
the Gradient Boosting Classifier algorithm, a powerful ensemble learning technique known for its
effectiveness in handling complex datasets and producing high-quality predictions.

Once the model is trained and validated, it will be integrated into a user-friendly application using
Streamlit, allowing users to input their medical data and obtain real-time predictions regarding their risk
of heart disease. This interactive platform aims to empower individuals to proactively monitor their
cardiovascular health and make informed decisions about lifestyle choices and medical interventions.

Through this project, we seek to contribute to the advancement of preventive healthcare by

leveraging machine learning technology to enhance early detection and intervention strategies for heart
disease, ultimately leading to improved patient outcomes and reduced healthcare burdens.

1.
2. Project Statement

Heart disease remains one of the leading causes of death globally, posing a significant public
health challenge. Despite advancements in medical science and technology, the timely identification and
prediction of heart disease risk factors remain elusive for many individuals. This lack of early detection
often results in delayed medical interventions, leading to adverse health outcomes and increased
healthcare costs. Additionally, the complexity and interplay of various risk factors make accurate
prediction challenging using conventional methods alone.

One of the primary challenges in addressing heart disease lies in the complexity of its etiology
and the multitude of factors contributing to its development. Traditional risk assessment methods often
rely on individual risk factors such as age, gender, and blood pressure, overlooking the intricate
relationships between these factors and their combined effect on cardiovascular health. This
oversimplified approach may lead to inaccurate risk assessments and missed opportunities for preventive
interventions.

Furthermore, the limited accessibility and usability of existing risk assessment tools hinder
widespread adoption and usage by individuals seeking to assess their heart disease risk. Many
conventional risk calculators require specialized medical knowledge to interpret results accurately,
presenting a barrier for individuals without medical expertise. This lack of accessibility prevents
proactive monitoring of cardiovascular health and delays timely interventions, exacerbating the burden of
heart disease on healthcare systems.

The absence of personalized risk assessment tools tailored to individual characteristics and
medical history further exacerbates the challenge of predicting heart disease accurately. Generic risk
models fail to account for the unique genetic predispositions, lifestyle factors, and comorbidities that may
influence an individual's susceptibility to heart disease. As a result, there is a pressing need for predictive
models capable of integrating diverse datasets and generating personalized risk assessments to guide
preventive strategies effectively.

Moreover, the dynamic nature of heart disease risk necessitates continuous monitoring and
adjustment of predictive models to accommodate evolving risk factors and medical knowledge. Static
risk assessment tools fail to adapt to changing health profiles and emerging risk factors, limiting their
long-term efficacy in predicting heart disease accurately. Therefore, there is a critical need for dynamic
predictive models capable of adapting to evolving healthcare landscapes and providing up-to-date risk
assessments.

2.
 Additionally, the lack of user-friendly interfaces and interactive platforms for accessing and
interpreting heart disease risk assessments impedes individuals' ability to engage actively in preventive
healthcare practices. Cumbersome interfaces and complex data visualization methods may discourage
individuals from utilizing risk assessment tools regularly, undermining efforts to promote proactive
cardiovascular health monitoring. Streamlining the user experience and enhancing accessibility are
essential components of effective heart disease prediction systems.

Furthermore, the reliance on traditional statistical methods for risk assessment may overlook
non-linear relationships and interactions among risk factors, limiting the predictive accuracy of existing
models. Machine learning algorithms offer a promising alternative by leveraging complex data patterns
and interactions to generate more accurate risk predictions. However, the successful integration of
machine learning techniques into heart disease prediction systems requires careful model selection,
feature engineering, and validation to ensure robust and reliable performance.

In summary, the overarching problem addressed by this project is the need for an accurate,
accessible, and personalized heart disease prediction system capable of leveraging machine learning
techniques to provide timely risk assessments and facilitate proactive cardiovascular health monitoring.
By addressing these challenges, we aim to empower individuals to take control of their heart health and
reduce the burden of heart disease on individuals and healthcare systems alike.

3.
3. Requirement Analysis

In order to develop a robust heart disease prediction system, a comprehensive analysis of

requirements is essential for the project.

● Data Collection:

1. Gather a comprehensive dataset containing relevant medical information such as age, gender,
blood pressure, cholesterol levels, and other pertinent indicators associated with heart disease.

2. Ensure the dataset is diverse and representative of the target population to avoid biases in the
predictive model.

● Data Preprocessing:

1. Cleanse the dataset by handling missing values, outliers, and inconsistencies.

2. Normalize or scale the features to ensure uniformity and improve model performance.

3. Perform exploratory data analysis (EDA) to gain insights into the distribution and
relationships among variables.

● Model Selection:

1. Choose an appropriate machine learning algorithm for heart disease prediction.

2. Assess various models and select the Gradient Boosting Classifier based on its performance
metrics, such as accuracy, precision, recall, and F1-score.

● Model Training and Validation:

1. Split the dataset into training and testing sets to evaluate the model's generalization
performance.

2. Utilize cross-validation techniques to optimize model hyperparameters and prevent

overfitting.

4.
 3. Evaluate the model's performance using appropriate evaluation metrics and validate its
effectiveness in accurately predicting heart disease outcomes.

● Integration with Streamlit Application:

1. Develop an interactive user interface using Streamlit to facilitate user interaction with the
predictive model.

2. Implement input forms to collect user data, ensuring confidentiality and privacy.

3. Integrate the trained model into the Streamlit application to enable real-time predictions of
heart disease risk based on user input.

● User Experience Enhancement:

1. Design a user-friendly interface with intuitive navigation and clear instructions for inputting
data.

2. Provide informative feedback to users regarding their predicted risk of heart disease and
potential actions for risk mitigation.

3. Incorporate visualization tools to present model predictions and insights in a comprehensible

manner.

● Deployment and Maintenance:

1. Deploy the Streamlit application on a suitable platform to make it accessible to users.

2. Regularly monitor the performance of the predictive model and update it as necessary with
new data or improvements.

3. Address any user feedback or issues to ensure the application remains reliable and effective in
supporting preventive healthcare initiatives.

By addressing these requirements, the heart disease prediction system can effectively fulfill its
objectives of early detection, risk assessment, and user empowerment in managing cardiovascular health.

5.
4. System Specification

● Zero Level DFD:

1. Data Collection: Kaggle.com, a well-known website for dataset access, is the source of the
data. Particularly, data regarding Heart Disease is collected.

2. Data Preprocessing: Preprocessing is done on the gathered data to make sure it is suitable
and of a high enough quality to train machine learning models. This covers managing missing
values, eliminating duplicates and carrying out feature scaling or data normalization.

3. Model Selection: During the model selection phase, the preprocessed data is prepared for
training different machine learning models. The Gradient Boosting Classifier model is
selected according to the performance and efficiency.

4. Training and Testing: Training and testing sets are created from the preprocessed data. The
training data is used to train the models, and the testing data is used to assess how well they
function. Each model's performance is assessed using accuracy as the evaluation metric.

5. Model Deployment: Streamlit is used to construct an interactive web application alongside

its cloud deployment features. It has an intuitive user interface. Upon selecting a particular
ailment, the software requests that the user input the necessary details for the prediction.

6.
 ● ER Diagram:

There are Four Entities in the above ER Diagram and following is the detailed explanation of the
above ER Diagram:

1. Data: Data is the dataset that includes the details about patients who may or may not have
heart disease. The dataset have various attributes such as:

● sex: gender of the patient

● age: age of the patient
● trestbps: resting blood pressure
● ср: serum cholesterol
● oldpeak: ST depression induced by exercise relative to rest
● exang: exercise-induced angina
● slope: the slope of the peak exercise ST segment
● thal: thalassemia
● fbs: fasting blood sugar > 120 mg/dl
● ca: number of major vessels colored by fluoroscopy
● chol: serum cholesterol
● restecg: resting electrocardiographic results
● thalach: maximum heart rate achieved
● target: target variable, indicating the presence of heart disease in the patient

7.
2. Model: The dataset is then used to train the model for predicting heart disease. The
GradientBoostingClassifier is used as the algorithm to train the model. This model is later
integrated to the user interface and is used to generate predictions by analyzing the user
inputs.

3. Streamlit UI: This model will be integrated into a user-friendly Streamlit app, providing an
intuitive interface for users to input relevant health data and receive accurate predictions
regarding their risk of heart disease. The Streamlit UI will offer a seamless experience,
empowering users to make informed decisions about their cardiovascular health with ease and
confidence.

4. User: The User with the help of Streamlit UI can submit their health parameters and with the
help of the Model will generate a prediction based upon the user input. The UI shows the
prediction to the user through the app. Thus the user will be able to know about his health
condition and will be able take the relevant precautions and medical consultation.

8.