International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 03 | Mar 2025
p-ISSN: 2395-0072
www.irjet.net
Cardiovascular Disease Prediction and Risk Assessment using Machine Learning Approaches Manpreet Hire1, Yasir Khan2, Zuber Shaikh3, Priyanshu Prajapati4 1Asst. Professor, Department of Data Science, Thakur College of Science and Commerce, Mumbai, India 2Student, Department of Data Science, Thakur College of Science and Commerce, Mumbai, India
3Student, Department of Data Science, Thakur College of Science and Commerce, Mumbai, India 4Machine Learning Engineer Intern, BNK Infotech Pvt. Ltd., Delhi, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In this research, we develop a machine
disease prognosis using a multifaceted healthcare dataset that includes patient demographics, health behaviors, and clinical history.
learning-based framework to predict patient-specific disease outcomes using a multifactorial healthcare dataset. The dataset encompasses a wide array of variables, including demographic details (such as age and sex), clinical indicators (general health condition, medical checkup frequency, physical activity), behavioral aspects (smoking history), and records of chronic illnesses like cardiovascular disease, skin cancer, diabetes, arthritis, and other cancers. Our exploratory data analysis revealed notable patterns in disease occurrence, highlighting how various health and lifestyle factors influence disease susceptibility. Furthermore, the dataset presented challenges in the form of incomplete data, underscoring the necessity for effective handling of missing values to ensure model robustness. The predictive models developed in this study offer valuable insights for identifying high-risk individuals and facilitating early interventions. The findings underline the potential of machine learning methodologies in enhancing clinical decision-making and advancing personalized healthcare.
The proposed model is developed and validated using a train-test split strategy to ensure robust and generalizable performance. A variety of evaluation metrics—including the confusion matrix, ROC curve, and Precision-Recall curve—are employed to provide a holistic assessment of the model’s effectiveness, particularly in managing class imbalance common in healthcare data. By integrating these methods, the study demonstrates the capacity of ML to significantly enhance disease risk assessment and inform clinical decision-making processes. The findings underscore the potential of predictive modeling to support healthcare professionals in identifying high-risk patients and tailoring early intervention strategies. Additionally, this research paves the way for future improvements through advanced data balancing techniques and the inclusion of supplementary clinical features, ultimately contributing to the broader adoption of data-driven, personalized healthcare solutions.
Key Words: Cardiovascular Disease Prediction, Machine Learning, Risk Assessment, Healthcare Data, Chronic Disease Forecasting, Clinical Analytics, Lifestyle Risk Factors, Predictive Modeling, Patient-Level Prognosis, Exploratory Data Analysis.
1.1 BACKGROUND Machine learning, particularly ensemble methods like gradient boosting, has proven to be highly effective in tackling these challenges. Algorithms such as XGBoost offer enhanced performance by iteratively reducing errors and handling imbalanced datasets commonly found in medical records. By integrating diverse patient-level data—ranging from demographic information to lifestyle factors—ML models can provide more precise and individualized risk assessments. This paradigm shift toward data-driven healthcare underscores the growing relevance of predictive analytics in clinical practice.
1. INTRODUCTION The growing burden of cardiovascular and other chronic diseases represents a significant challenge for modern healthcare systems. These conditions are influenced by a complex combination of demographic, behavioral, and clinical factors, necessitating accurate and early risk prediction to improve patient outcomes and reduce healthcare costs. Conventional risk scoring systems often fail to capture the intricate relationships embedded within large-scale patient data, creating a demand for more advanced and flexible predictive tools. In this context, machine learning (ML) has emerged as a transformative approach in healthcare analytics, enabling the discovery of hidden patterns that traditional statistical models may overlook. This study focuses on leveraging ML, particularly gradient boosting algorithms such as XGBoost, to forecast
© 2025, IRJET
|
Impact Factor value: 8.315
1.2 TOOLS & TECHNOLOGIES This study primarily leverages Python, a versatile and widely adopted programming language in the field of data science and machine learning, due to its extensive libraries and ease of integration in healthcare analytics workflows. The core machine learning framework utilized in this
|
ISO 9001:2008 Certified Journal
|
Page 700