Deep Learning-Based Classification of Lung Cancer

Deep Learning-Based Classification of Lung Cancer

J.Noor Ahamed1, Jeshika J2

1Associate professor, Department of Computer Applications, Nehru College of Management, Coimbatore, Tamil Nadu, India.

ncmnoorahamed@gmail.com

2Student of II MCA, Department of Computer Applications, Nehru College of Management, Coimbatore, Tamil Nadu, India.

jeshikajustin@gmail.com

Abstract

Lung cancer continues to be one of the deadliest diseases globally, largely due to late diagnosis and restricted access to advanced screening technologies. This research introduces a deep learning-based method for the automated classification of lung cancer utilizing the Vision Transformer (ViT) model. The suggested system employs the ViT architecture to analyze CT scan and X-ray images, effectively capturing both local and global spatial relationships through self-attention mechanisms. The dataset includes images of adenocarcinoma, large-cell carcinoma, squamous-cell carcinoma, and healthy lung tissue. Images undergo preprocessing through resizing, normalization, and augmentation to improve model robustness. The ViT model is trained and assessed using metrics such as accuracy, precision, recall, and F1-score, and is compared with conventional convolutional neural network (CNN) models. The experimental findings indicate that the ViT-based model achieves enhanced classification performance, facilitating more reliable and early detection of lung cancer. The system is implemented as a Flask-based web application, offering healthcare professionals a real-time diagnostic interface that enables the upload and automated analysis of medical images. This study underscores the potential of Vision Transformers in clinical diagnostics and contributes to the advancement of effective, AI-assisted tools for the early detection of lung cancer.

Keywords

Deep Learning; Vision Transformer (ViT); Lung Cancer Classification; Medical Image Analysis; Convolutional Neural Networks (CNN); Flask Framework; Image Preprocessing; Computer-Aided Diagnosis (CAD); CT Scan; X-ray Imaging