Advanced AI Techniques for Interstitial Lung Disease Detection

Project Overview

Pulmonary diseases, including asthma, chronic bronchitis, and pulmonary fibrosis, are becoming increasingly prevalent worldwide. These conditions are often aggravated by environmental pollution and unhealthy lifestyle factors, making early and accurate diagnosis essential for effective treatment and improved patient outcomes. Recent advances in Artificial Intelligence (AI) have opened new possibilities for enhancing medical diagnostic processes. In this context, this project focuses on multi-class classification of pulmonary fibrosis using CT scan images and Transformer-based deep learning models ().

Proposed AI Approach

• CNN Baseline Model: Implemented a Convolutional Neural Network (CNN) to extract spatial features from CT scan images and evaluate initial classification performance.
• VGG16 Architecture: Applied the VGG16 deep learning model on CT scan data to enhance feature extraction and improve classification accuracy.
• Vision Transformers (ViT): Integrated Vision Transformers to capture global contextual relationships within medical images and detect complex fibrosis patterns.
• Performance Comparison: Compared CNN, VGG16, and ViT models using different metrics to determine the most effective architecture.

Data & Methodology