Radiomics applied to lung cancer: a review

Abstract: Lung cancers exhibit strong phenotypic differences that can be visualized noninvasively by medical imaging. Radiomics, a concept introduced in 2012, refers to the comprehensive quantification of tumor phenotypes by applying a large number of quantitative image features (watch the animation: https://youtu.be/Tq980GEVP0Y and the website www.radiomics.org). Here, we review the literature related to radiomics for lung cancer. We found 11 papers related to computed tomography (CT) radiomics, 3 to radiomics or texture analysis with positron emission tomography (PET) and 8 relating to PET/CT radiomics. There are two main applications of radiomics, the classification of lung nodules (diagnostic) or prognostication of established lung cancer (theragnostic). There are quite a few methodological issues in most of the reviewed papers. Only 5 studies, out of the 22, were externally validated. Overall, it is clear that radiomics offers great potential in improving diagnosis and patient stratification in lung cancer. It may also have a real clinical impact, as imaging is routinely used in clinical practice, providing an unprecedented opportunity to improve decision support in lung cancer treatment at low cost.