TB-Analyzer

Project Overview 🚀

TB-Analyzer is an AI-based web application designed to detect tuberculosis (TB) using chest X-rays. Built with TensorFlow.js, it provides a platform for quick and accessible TB screening.

Features ✨

• AI-Powered Detection: Utilizes machine learning models trained on public datasets.
• User-Friendly Interface: Web-based application accessible through any modern browser.
• Instant Results: Provides immediate analysis and feedback on uploaded X-ray images.

Technical Implementation 🛠️

• Model Training: Used TensorFlow to train the model using two public image datasets.
• Frontend: Developed using JavaScript and TensorFlow.js for running models in the browser.
• Deployment: Hosted on a scalable platform to handle multiple simultaneous users.

Challenges & Solutions 🤝

• Challenge: Handling large model sizes for efficient browser execution.
  • Solution: Optimized the model and utilized TensorFlow.js capabilities to run models efficiently in the browser.
• Challenge: Ensuring accuracy with limited dataset availability.
  • Solution: Combined multiple public datasets and implemented data augmentation techniques to improve model performance.

License 📄

This project is open-source and available under the MIT License.

• JavaScript
• TensorFlow.js
• AI/ML