The Automated Breast Cancer Predictor uses AI to classify tumors with high accuracy, providing quick predictions and shareable reports for better healthcare decisions.
🎗️ AI-Based Breast Cancer Predictor Web App
🌐 Live Demo(Render Deploy): https://automated-breast-cancer-predictor-com.onrender.com
This is an AI-powered web application that predicts the likelihood of breast cancer (benign or malignant) using a Logistic Regression model built with scikit-learn. Designed for early detection, it helps in making clinical decisions with 86%+ accuracy.
Key Features:
Trained ML model (Logistic Regression on 8 highly correlated features)
Real-time prediction based on user input
PDF report generation + Email sharing
Role-based access (Doctor & other Resercher)
Animated and responsive UI with particle effects
Live deployment on Render
🔧 Tech Stack:
Frontend: HTML, CSS, JavaScript, AOS, Particles.js, Chart.js
Backend: Python, Flask, scikit-learn, ReportLab
Database: SQLite
Deployment: Render
ML Algorithm: Logistic Regression (scikit-learn)
Dataset: Breast Cancer Wisconsin Dataset View on Kaggle
⏱️ Progress During 48-Hour Hackathon In the 48-hour hackathon, we built the Automated Breast Cancer Predictor, focusing on real-world impact and accessibility. On Day 1, we ideated around healthcare challenges and finalized the Breast Cancer Wisconsin dataset. We cleaned the data, selected the 8 most significant features, and trained a Logistic Regression model using scikit-learn, achieving over 95% accuracy. Next, we built a Flask backend to process patient data, perform predictions, generate PDF reports, and send them via email. Simultaneously, we developed a clean, responsive frontend UI with HTML, CSS, JavaScript, AOS animations, and Particles.js for visual appeal. We integrated Chart.js for interactive data visualization and implemented user authentication with login/signup and role-based access using SQLite. Finally, we deployed the full-stack application on Render and performed final testing.