Chest Disease Detection In X-Ray Images Using Deep Learning Classification Method

TL;DR

This study tackles autonomous, interpretable chest-disease diagnosis from X-ray images across four classes: COVID-19, pneumonia, TB, and normal. It leverages transfer learning with multiple pretrained CNNs and uses Grad-CAM to provide visual explanations of predictions, achieving high accuracy and robust metrics on a large, diverse dataset (57,111 images). Among models, ResNet50 offers the strongest overall performance, with Grad-CAM aiding interpretability and revealing areas concordant with expert assessment in several cases, though some attention regions extend outside the lungs indicating dataset-bias concerns. The work highlights practical implications for deployment in resource-limited settings, emphasizes the role of explainability in clinical trust, and outlines future directions including dataset expansion, preprocessing to reduce bias, and integration with language models to support clinicians.

Abstract

In this work, we investigate the performance across multiple classification models to classify chest X-ray images into four categories of COVID-19, pneumonia, tuberculosis (TB), and normal cases. We leveraged transfer learning techniques with state-of-the-art pre-trained Convolutional Neural Networks (CNNs) models. We fine-tuned these pre-trained architectures on a labeled medical x-ray images. The initial results are promising with high accuracy and strong performance in key classification metrics such as precision, recall, and F1 score. We applied Gradient-weighted Class Activation Mapping (Grad-CAM) for model interpretability to provide visual explanations for classification decisions, improving trust and transparency in clinical applications.

Figures (18)

• Figure 1: Chest X-ray images
• Figure 2: CNN Architecture
• Figure 4: Training Accuracy over 10 Epochs
• Figure 5: Training Loss over 10 Epochs
• Figure 6: Validation Accuracy over 10 Epochs