New Perspectives for Estimating Body Composition From Computed Tomography: Clothing Associated Artifacts.

As the value of clinical imaging is expanded through retrospective analyses, it is imperative that all efforts are made to optimize validity. Such considerations for retrospective designs should prioritize factors like naturalistic conditions for observations and measurement replicability, while avoiding sample biases and reliance on strict clinical timelines. Valid methodological approaches are immanent for successful translation from retrospective observational designs into prospective pragmatic research with actionable potential.