Abstract
Personalized treatment is one of the major challenges in oncology today. This is made all the more complex by the heterogeneity of the cells within a tumor and their evolution over time.
Radiomics is a new image processing technique based on bioinformatics and biostatistics techniques, designed to address the dual challenge of molecular profiling and tumor heterogeneity. It is based on the calculation of a very large number of parameters from the image (several hundred to several thousand) in order to bring out new relevant parameters that may not have been previously identified. Its objectives are to identify imaging profiles associated with a genetic or molecular profile (also known as radiogenomics), or to predict therapeutic efficacy in order to guide personalized treatment choices. The challenges, principles and limits of radiomics will be discussed during the lecture.
