Assistance to lung cancer detection on histological images using Convolutional Neural Networks

Cancer is one of the leading causes of death worldwide, accounting for nearly 10 million deaths in 2020, according to the World Health Organization. Lung cancer is the one with the highest mortality. Its detection in early stages is crucial to prevent its spread, known as metastasis. During these last decades, the interpretation of medical imaging (computed tomography, magnetic resonance imaging, echography, etc.) has been performed by human experts. However, recently, these analyses have begun to benefit from the help provided by AI-based systems. These systems are still not able to replace the experts, but they can help as diagnostic assistance. In this sense, this study aims to design a tool for detecting lung cancer through the analysis of histopathologic images using Deep Learning models. Moreover, a whole labelled dataset has been elaborated with the clinical data of 21 patients provided by the Hospital Universitario Torrecárdenas of Almería.