Multi-Class Classification of Chest X-ray Images Using a Residual Capsule Network

This study introduces a novel deep learning network, ResCapsNet (Residual Capsule Network), for classifying patients as having COVID-19, tuberculosis (TB), or pneumonia using chest X-ray images. These networks address limitations of existing methods and achieve high accuracy, showcasing its potential as a valuable tool for automated diagnosis. Additionally, a user-friendly interface was developed for easy access by both healthcare professionals and the general public. Users can upload their chest X-rays (or those of a concerned individual) after signing up for the secure website and receive a predicted diagnosis. This deep learning system analyzes and classifies the uploaded chest X-ray images using a hybrid deep learning model, namely, ResCapsNet. The uploaded image undergoes preprocessing (such as resizing, normalization, etc.) to ensure consistency. The preprocessed image is then fed into the hybrid model, which combines improved residual modules for feature extraction and capsule layers for final classification.