Deep Learning Based Lung Cancer Detection, Classification and Its Localization

Lung cancer remains one of the deadliest causes for millions of deaths each year. In India, the increasing number of lung cancer cases underscores the urgent need for early and accurate detection. This study presents a deep learning based approach for the detection, classification, and localization of lung cancer using CT scan images. Leveraging Convolutional Neural Networks (CNN) and advanced image segmentation models like U-Net, the proposed system aims to assist radiologists in diagnosing lung cancer with higher precision and speed. The system automatically identifies cancerous regions, classifies cancer into benign or malignant categories, and localizes affected areas within the lungs through heatmap visualization and Grad – CAM explainability. Experimental results demonstrate an impressive accuracy rate of 93.2%, with strong precision and recall values, validating the robustness of the model. By integrating machine learning and deep learning, this research enhances diagnostic accuracy, reduces manual effort, and supports timely medical intervention. The proposed system signifies a step forward in developing reliable, interpretable, and efficient AI-assisted diagnostic tools for lung cancer detection, ultimately contributing to improved patient outcomes and saving lives. CCS CONCEPTS · Computing methodologies → Deep learning; Neural networks; Image processing;