Lung Cancer Subtype Classification Using Deep Learning Models

Abstract

Lung cancer is one of the deadliest cancers, primarily due to late-stage diagnosis. Traditional histopathological examination is time-consuming, labor-intensive, and prone to human error, highlighting the need for automated solutions. This project leverages deep learning to classify lung cancer using the LC25000 dataset, which includes histopathological images of lung adenocarcinoma, squamous cell carcinoma, and benign tissue. The model utilizes transfer learning and data augmentation to enhance accuracy while minimizing overfitting, ensuring reliable predictions. By automating image analysis, this AI-driven approach improves diagnostic efficiency, reduces the workload of pathologists, and minimizes diagnostic errors. The integration of deep learning into lung cancer diagnosis enables faster and more accurate classification, assisting healthcare professionals in making informed decisions. With the growing adoption of AI in healthcare, this project demonstrates the potential of artificial intelligence in improving early detection, treatment planning, and overall patient outcomes in lung cancer diagnostics.