CNN Chest X-Ray Classification

Authors: Mateusz Andrzejewski, Wiktor Siepka

This project focuses on using Convolutional Neural Networks (CNNs) to classify chest X-ray images for detecting tuberculosis abnormalities. This project was developed as part of the Neural Networks university course.

Features

• Dataset: Utilizes the Pulmonary Chest X-Ray Abnormalities dataset from Kaggle, consisting of healthy and sick patient X-rays.
• Model: Implements a custom CNN architecture with 4 convolutional layers, Max Pooling, and Dropout to prevent overfitting.
• Training: Uses the Adam optimizer and Early Stopping; trained for 50 epochs with a batch size of 32.
• Evaluation: Achieved 83% accuracy on the test set.