Visual analytics: design, measures, classic example

An estimated 40% of all cancer cures have been attributed to radiotherapy treatments, an outcome made possible by treatment guidelines optimized for tumor type, size, and shape. Unfortunately, such tumor-based guidelines often result in a wide variation in treatment responses, despite the patients having similar clinical profiles. Recent research suggests that this wide variation in responses could be the result of molecular heterogeneity across patients, a well-known characteristic of most cancers. This chapter discusses how advanced methods from visual analytics could enable (1) the quantitative modeling of molecular heterogeneity to identify patient subgroups with distinct molecular profiles and outcomes, and (2) the visual modeling of patient subgroups to enable clinicians to infer mechanisms that precipitate adverse outcomes in each subgroup, with the goal of designing precision radiotherapy treatments.