Lung Cancer Detection Systems Applied to Medical Images: A State-of-the-Art Survey

Abstract

Lung cancer represents a significant global health challenge, transcending demographic boundaries of age, gender, and ethnicity. Timely detection stands as a pivotal factor for enhancing both survival rates and post-diagnosis quality of life. Artificial intelligence (AI) emerges as a transformative force with the potential to substantially enhance the accuracy and efficiency of Computer-Aided Diagnosis (CAD) systems for lung cancer. Despite the burgeoning interest, a notable gap persists in the literature concerning comprehensive reviews that delve into the intricate design and architectural facets of these systems. While existing reviews furnish valuable insights into result summaries and model attributes, a glaring absence prevails in offering a reliable roadmap to guide researchers towards optimal research directions. Addressing this gap in automated lung cancer detection within medical imaging, this survey adopts a focused approach, specifically targeting innovative models tailored solely for medical image analysis. The survey endeavors to meticulously scrutinize and merge knowledge pertaining to both the architectural components and intended functionalities of these models. In adherence to PRISMA guidelines, this survey systematically incorporates and analyzes 119 original articles spanning the years 2019–2023 sourced from Scopus and WoS-indexed repositories. The survey is underpinned by three primary areas of inquiry: the application of AI within CAD systems, the intricacies of model architectural designs, and comparative analyses of the latest advancements in lung cancer detection systems. To ensure coherence and depth in analysis, the surveyed methodologies are categorically classified into seven distinct groups based on their foundational models. Furthermore, the survey conducts a rigorous review of references and discerns trend observations concerning model designs and associated tasks. Beyond synthesizing existing knowledge, this survey serves as a guide that highlights potential avenues for further research within this critical domain. By providing comprehensive insights and facilitating informed decision-making, this survey aims to contribute to the body of knowledge in the study of automated lung cancer detection and propel advancements in the field.