CNNs-based tool for lung cancer prediction from CT scans

Lung cancer is the most deadly type of cancer worldwide, an important part of this mortality is due to the fact that health professionals do not have enough tools that allow them to make an adequate and in-time diagnosis. This paper explores the development and validation process of a tool that uses artificial intelligence to aid in the detection of cancer in lung CT scans. The generated model is able to reach an effectiveness of 76% in the classification of 4 different types of classes, being large cell carcinomas the most difficult type of cancer for the model to detect. Finally, it proposes the use of pre-trained neural networks and their integration in scalable computer systems to facilitate the widespread use of this type of tools in the medical field.