Globally, lung cancer accounts for the most cancer deaths in both sexes combined. It is believed to develop slowly through progressive accumulation of genetic mutations, hence the disease allows time for diagnosis and curative surgical treatment. Five year survival rates for non-small-cell lung carcinoma (NSCLC) can range from 57-61% when detected in the early stages of disease. This is compared with a survival rate of approximately 6% once distant metastases are present. However, disease diagnosis typically occurs when it has progressed to an advanced stage when patients present with signs and symptoms. Therefore, technologies capable of asymptomatic disease detection will significantly impact lung cancer specific mortality. Metabolomic profiling of cancer measures compounds produced as a result of cellular activity including volatile organic compounds (VOCs) in exhaled breath. Infrared spectroscopy is a proven technique for breath analysis that can measure chemical concentrations in the parts per trillion range for certain VOCs. When coupled with machine learning techniques, this has the potential to be a novel approach for disease detection using exhaled breath.
