Comprehensive statistical analysis of radiomics data by RadAR, a new package for R

The high-dimensional quantitative analysis of biomedical images, referred to as Radiomics, is emerging as a promising approach to facilitate clinical decisions and improve patient stratification. A typical radiomics workflow can be divided into two main analytical processes, namely data generation and data analysis. The first process includes image acquisition, segmentation and feature extraction. The second process involves the statistical analysis of high-dimensional datasets that may include hundreds of observations (i.e., feature values) per patient and the integration of these data with other available clinical/molecular information. While procedures for primary radiomics analysis (data generation) has been sufficiently established during the last years, secondary analysis remains challenging, representing a critical step of the entire radiomics workflow. Here we present Radiomics Analysis with R (RadAR), a novel freely available computational tool to perform comprehensive secondary analysis of radiomics features.