Lung Cancer Detection using Weakly-Supervised Learning for: Leveraging the ChestX-ray8 Dataset for Automated Classification

Timely detection of lung diseases is crucial for effective treatment and improved patient outcomes. This research advances medical image analysis by applying weakly-supervised learning techniques to chest X-ray images, using the “ChestX-ray8” dataset with 108,948 frontal-view X-ray images from 32,717 patients. Our automated lung cancer detection methodology employs a deep learning model, emphasizing weakly-supervised learning to address challenges posed by limited annotations. The model integrates a pre-trained MobileNet for feature extraction and custom layers for disease classification. Key stages, including data handling, exploration, visualization, image preprocessing, model compilation, training, evaluation, and optional continuation, are outlined. Evaluation metrics, such as ROC curves, assess the model’s discriminatory power for various lung diseases. The code follows best practices, and the final trained model is serialized for future use. This research contributes to enhancing computer-aided diagnosis systems for lung diseases, providing a promising avenue for further exploration in medical image analysis.