Inspiration
Cardiovascular diseases are a major healthcare concern. We wanted to combine AI and 3D visualization to make coronary artery disease risk predictions easier to explore and understand.
What It Does
CardioScope3D is a full-stack web application that predicts disease-related outcomes for CAD, LAD, LCX, and RCA using machine learning models. It also provides interactive 3D heart visualization to explore coronary artery anatomy.
How We Built It
- Frontend: React.js, Vite, Three.js
- Backend: Python, FastAPI
- Machine Learning: Scikit-learn, XGBoost
- Data Processing: Pandas, NumPy
- Explainable AI: SHAP
The application connects patient input, machine learning inference, probability estimates, and interactive visualization through REST APIs.
Challenges We Faced
Integrating multiple ML models, maintaining consistent input features, managing model dependencies, connecting the frontend with the backend, and presenting predictions responsibly were our main challenges.
What We Learned
We gained practical experience in full-stack development, ML model integration, API development, explainable AI, and 3D visualization.
Future Scope
We plan to improve model explainability, enhance coronary artery visualization, evaluate model performance, and communicate prediction uncertainty more effectively.
Disclaimer: This project is intended for educational and research purposes only. Its predictions are not medical diagnoses and require independent clinical validation before clinical use.
Log in or sign up for Devpost to join the conversation.