The epistemology of imaging procedures and reporting

The good epistemic standing of imaging procedures is in turn epistemically complex, as it involves three levels of normative assessment. First, it depends on rules and conventions about how to read images (the “semantics” of imaging); second, on how these rules and conventions are applied by the individual reader (either radiologist or nuclear medicine expert) in the specific case (usually called “reading”); and third, on how the reading is conveyed in a written report (“reporting”). That imaging involves conventions may appear controversial (there is an element of arbitrariness in conventions, and what could be arbitrary in the fact that a focal hot spot in a [18F]-FDG PET image of the lung should be read as cancer?). We will consider this in more detail below, with particular emphasis on reporting.