AI Powered Pneumonia Detection

Abstract

Pneumonia is a significant health concern, particularly in developing regions where timely and accurate diagnosis can be challenging. This project focuses on building an AI-based Pneumonia Detection System using ResNet152v2, a deep learning model, to detect pneumonia from chest X-ray images. The system leverages TensorFlow for model training and Flask to create an accessible web application where users can upload images for real-time diagnosis. The dataset used in this project consists of X-ray images categorized as either ’normal’ or ’pneumonia.’ To address the class imbalance in the dataset, techniques such as data augmentation and class weighting were employed. To enhance functionality and ensure a seamless user experience, MongoDB was integrated as the database for storing user data, diagnostic results, and uploaded images securely. This integration allows users to access their diagnosis history and facilitates data analysis for further research. A modular architecture was adopted, separating the user interface, backend API, database, and deep learning model to ensure maintainability, scalability, and efficient data management. This project demonstrates the potential of AI to improve healthcare outcomes by providing an accessible and reliable diagnostic tool.