Pulmonary Cancer Prediction Using Machine Learning

Abstract

Lung cancer is one of the leading causes of cancer-related deaths worldwide, and its early detection plays a crucial role in improving survival rates. Traditional diagnostic methods rely heavily on radiologists manually analysing CT scan images, which can be time-consuming, subjective, and prone to human error. To address these challenges, this project introduces an AI-powered Pulmonary Cancer Detection System that utilizes Convolutional Neural Networks (CNNs) to automatically classify lung CT scans into four categories: Adenocarcinoma, Large Cell Carcinoma, Squamous Cell Carcinoma, and Normal.
The system is trained on a large dataset of medical CT scans, ensuring high accuracy through the application of data augmentation, batch normalization, and dropout layers to prevent overfitting. To make this technology accessible, a Streamlit-based web application has been developed, allowing users to upload CT scan images and receive real-time predictions with confidence scores and probability visualizations. The model is designed to assist medical professionals, researchers, and healthcare institutions by providing a fast, reliable, and automated approach to lung cancer detection. By leveraging deep learning techniques, this system reduces manual diagnosis time, enhances early detection accuracy, and aids in clinical decision-making. The integration of interactive visualizations and probability scores ensures transparency in the model’s predictions, helping radiologists interpret results effectively. This project not only demonstrates the potential of AI in medical imaging but also serves as a stepping stone for future advancements in computer-aided diagnosis (CAD) systems. Further improvements, such as integrating explainable AI (Grad-CAM) and expanding the dataset with diverse medical scans, could enhance the system’s reliability and real-world applicability.