Table 2.

Statistical Metrics Used in the Claim Statements

Example	Rationale	Statistical Metric	Interpretation
ADC in solid tumors	Cross-sectional claim with negligible bias	Within-subject (or within-tumor) Standard Deviation (wSD) × (95% confidence factor)	The variability seen in multiple measurements on a subject when no biologic change has occurred and the same imaging procedures are used for all measurements.
Change in Perc 15 for monitoring emphysema	Longitudinal claim; Same imaging procedures at 2 time points	Repeatability Coefficient (RC)	The difference between any two measurements on a case is expected to fall between −RC and +RC for 95% of replicated measurements. It represents the minimum detectable difference, with 95% confidence.
% change in CT tumor volume	Longitudinal claim; Different imaging procedures at 2 time points	Reproducibility Coefficient (RDC) is used in the claim statement. The Total Deviation Index with 95% coverage for change (TDI95%) is used for testing conformance.	The RDC is a measure of precision that is used when imaging procedures differ at the two time points. It has a similar interpretation as the RC: it is the minimum detectable difference, with 95% confidence. Since it can be measured directly from clinical studies, it is used in the claim statement. When testing conformance, the bias of the imaging procedures at the two time points will likely differ and this must be accounted for. Thus, the TDI is used for conformance testing. It includes components of both precision and bias. 95% of the differences between the measurements and their true value are <TDI95%. The TDI cannot be measured directly; rather, an actor’s bias and precision are estimated in separate studies. See section on Trade-off between Precision and Bias