Statistics of Referencing Susceptibility Maps in the Context of Clinical QSM Studies

Motivation: In QSM, there is no well-established susceptibility baseline . This can be determined a-posteriori by referencing to a specific tissue but this may impact statistics in clinical studies. Goal(s): To derive an expression for a t-test under referencing, and to investigate the effect of commonly used reference regions on a temporal lobe epilepsy study. Approach: Reference regions were compared: three anatomical structures and three derived from global thresholds. Changes in covariances, t-test results, and regional susceptibility distributions are presented. Results: Referencing to small regions has a bigger impact on statistical analyses than large references. Reference regions should have a low variance across groups. Impact: Referencing QSM susceptibility values is essential, but highly contested in practice, particularly in clinical applications. We clarify the statistical theory, and investigate the impact of referencing susceptibility measurements to different regions to facilitate practical implementation and clinical applications.