Abstract
Computed tomography (CT) is one of the widely used imaging modalities to monitor the development and progression of cancers. The quantitative CT imaging for personalized cancer medicine become increasingly attractive field. The underlying hypothesis of this research area is that the advanced computational approaches discover imaging biomakers associated with cancer probabilities, clinicopathological prognostic factors, and gene-expression levels from large amounts of image-based features. If this hypothesis is proven through external and independent validation cohorts of patients, we can noninvasively infer biological characteristics of diseases, possibly representing cancer probability and prognostic information, from the quantitative CT imaging. The purpose is to develop computer-aided detection/diagnosis (CADe/CADx) systems based on the multidisciplinary computational anatomy models which support clinicians to detect early-stage cancers and decide risk-adaptive treatments. These CADe/CADx systems may have a large impact as imaging is routinely used in clinical practice, in all stages of diagnoses and treatment, providing an unprecedented opportunity to improve medical decision-support.
