Adequacy of the patient-specific CTV robust margin

Determine whether a patient-specific, direction-resolved CTV robust margin derived as one standard deviation of the propagated-contour ensemble is sufficient for target coverage in head and neck online adaptive proton therapy.

Background

The uncertainty-guided digital twin framework derives direction-resolved CTV robust margins from the dispersion of propagated CTV contours across an ensemble of predicted treatment-day CTs. In the ten-patient feasibility study, these margins ranged from 1.1 to 3.0 mm and were used as setup uncertainties in robust optimization.

The authors explicitly state that the available comparison cannot establish whether margins of this size are sufficient, because the clinical offline replans used for comparison had their own clinical margins. Establishing margin sufficiency would require evaluation against an appropriate independent standard, likely involving broader patient cohorts and treatment-day anatomy beyond the single replanning QACT used here.