Real World Lung Cancer Screening Challenges

Lung cancer screening has the potential to save millions of lives. Sufficient evidence has now been provided to support the implementation of lung cancer screening worldwide. Implementation, however, is complex and presents new challenges which mandate further scientific research. Key findings from this thesis could be used to help address some of the challenges associated with the implementation of a lung cancer screening program. Part I (chapter 1 and 2) explored existing literature and international expert opinions to comprehensively review and highlight challenges associated with the implementation of global lung cancer screening programs. Part II (chapters 3 and 4) investigated the association between Influenza season and incidence of baseline detected lung nodules using data collected during the ImaLife general population study, and also resolution of new nodules detected in the NELSON study which included participants at high-risk of developing lung cancer. Part III (chapters 5-7) introduced the concept of using AI as a first-reader to rule out negative baseline CT scans in order to achieve significant reduction in radiologist CT-reading workload. This concept was explored using data from MLCS and then validated in the UKLS dataset with histological outcomes providing the gold standard. Finally, supporting evidence on the implementation of AI as a first reader was provided using data from the ongoing 4ITLR study. In this chapter, a summary of main findings per chapter will be presented followed by general discussion, future perspectives and conclusions.