More than lung cancer: Automated analysis of low-dose screening CT scans

Abstract

Smoking is a major health care problem and is projected to cause over 8 million deaths per year worldwide in the coming decades. To reduce lung cancer mortality in heavy smokers, several randomized screening trials were initiated in the past years using screening with low-dose Computed Tomography (CT). Recently, the National Lung Screening Trial (NLST), which was performed in the United States of America and compared about 25,000 participants screened with CT against 25,000 participants screened with conventional chest radiography, reported an overall lung cancer mortality reduction over 20% in the CT arm. Interestingly, this is also the first cancer screening that demonstrated an increase in overall survival, which raises high expectations. Despite cost-effectiveness is still awaited, major authorities have already released practical guidelines that recommend CT-based screening for lung cancer. Smoking-induced mortality and morbidity is not only due to lung cancer, and comparable numbers of heavy smokers die from cardiovascular disease (CVD), followed by chronic obstructive pulmonary disease (COPD). Interestingly, the screening test used for lung cancer screening (ie. CT) is also capable of visualizing several other pulmonary and cardiovascular diseases. This raises the question why one would only look at lung cancer. Quantitative assessment of disease components might prove useful in the identification of COPD in CT-based lung cancer screening setting. Also, cardiovascular risk may additionally be evaluated in lung cancer screening participants by quantitative analysis of vascular calcifications. If screening CT scans would enable early detection of these diseases, morbidity and mortality from CVD and COPD might be prevented, which may also enhance the cost-effectiveness of CT-based screening in heavy smokers. This thesis describes the automated analysis of low-dose CT scans in a lung cancer screening setting. It shows the optimization of quantitative analysis in expiratory CT, and reports the diagnostic and prognostic value of CT for evaluation of CVD and COPD. It shows that lung cancer screening CT contains additional information on diseases other than lung cancer, and can be used to identify subjects with early stages of COPD and CVD.