Lung Cancer Detection Using Transfer Learning

Abstract

Since lung cancer is still one of the most common and deadly types of cancer in the world, precise and effective screening techniques are desperately needed. Convolutional neural networks (CNNs), in particular, are deep learning algorithms that have demonstrated tremendous potential in a variety of medical image processing applications in recent years. Deep learning's area of transfer learning uses massive datasets of pre-trained networks to adapt models to domains with sparse labeled data, hence increasing the models' efficacy even further. This research uses deep learning and transfer learning approaches to give an extensive review and analysis of current developments in lung cancer detection. We talk about the difficulties of diagnosing lung cancer, such as interpreting complicated medical imaging, the disparity in class, and the scarcity of available data.In the area of lung cancer detection, we also include a summary of frequently used datasets, pre-processing methods, model topologies, and assessment measures. By means of a critical analysis of extant literature, we want to accentuate the merits and demerits of present methodologies and pinpoint prospective directions for further investigation. In the end, we want to support further efforts to provide precise, scalable, and clinically meaningful solutions for lung cancer early detection and treatment.