Dosimetric influence of deformable image registration uncertainties on propagated structures for online daily adaptive proton therapy of lung cancer patients

AbstractPurpose A major burden of introducing an online daily adaptive proton therapy (DAPT) workflow is the time and resources needed to correct propagated contours. In this study, we evaluated dosimetric impact neglecting correction contours in a DAPT workflow. Material methods For five NSCLC patients with nine repeated deep-inspiration breath-hold CTs, plans were optimised on planning CT deliver 60 Gy-RBE 30 fractions. All CTs registered six different clinically used deformable image registration (DIR) algorithms corresponding CT. Structures rigidly each DIR algorithm reference structures contoured uncorrected, (propagated doses) (ideal doses), non-adapted doses recalculated all CTs. Results Due anatomical changes occurring during therapy, clinical target volume (CTV) coverage reduces average by 9.7% (V95) compared ideal doses. propagated doses, CTV was always restored (average differences V95 < 1% doses). Hotspots reduced any approach. Conclusion presented here, benefit shown, even if optimisation based some residual uncertainties. However, careful (offline) structure review necessary corrections can be included offline adaption.