Design and Development of an AI-Enhanced Health Diagnostic System for Symptom-Based Prediction of Medical Conditions and Remedy Suggestions

Tuberculosis (TB) remains a significant global health challenge, particularly in resource-constrained settings. Early and accurate diagnosis is critical to improving patient outcomes and reducing transmission. This project presents an automated TB detection system powered by a Convolutional Neural Network (CNN) to analyze chest X-ray images and classify them as TB-positive or TB-negative. Leveraging transfer learning with a pre-trained deep learning model, the system achieves high accuracy while maintaining computational efficiency.

The application integrates a user-friendly web interface developed using Flask, enabling users to upload chest X-ray images for analysis. Upon processing, the system provides predictions alongside confidence scores. Additionally, the results are stored in a relational database using SQLite, allowing users to access a comprehensive history of predictions for monitoring and evaluation.

This project demonstrates the potential of deep learning in augmenting traditional diagnostic methods by offering an accessible, scalable, and reliable tool for TB detection. By combining medical imaging, artificial intelligence, and database management, the system aims to support healthcare providers in early TB diagnosis, particularly in low-resource environments where diagnostic tools may be limited.