Click here to

Session: Multi-Disciplinary General ePoster Viewing [Return to Session]

Evaluation of Maximum Permissible OAR Delineation Error That Avoids Clinical Impact Pancreatic Cancer Cases

Fergany Badry1*, MNH Rashad1, V Leandro Alves1, J Siebers1, (1) University of Virginia, Charlottesville, VA,


PO-GePV-M-206 (Sunday, 7/10/2022)   [Eastern Time (GMT-4)]

ePoster Forums

Purpose: To evaluate the maximum patient-specific OAR delineation error permissible that ensures that the OAR dose remains below a dose or dose-volume objective-based threshold for pancreatic cancer cases. This knowledge enables an efficient delineation review workflow.

Methods: For clinical fifteen pancreatic cases, OAR delineations were expanded by uniform margins (M) ranging from 4-20mm. The patient prescription doses ranged from 38-60 Gy. Percent dose-volumes metrics (Dx%[Gy]) of the non-expanded OARs were converted to absolute dose-volume metrics (Dxcc[Gy]) for the expanded OAR (eOAR) evaluations. Each eOAR dose-volume value was compared with a dose-volume-tolerance threshold (Dxcc[Gy]
Results: With f=1, 20mm delineation errors are permissible for all Kidney_left, Liver, and BowelBag cases; 13/15 Kidney_Right and SpinalCanal; 11/15 Stomach; and 12/15 Duodenum. For f=0.5, 20mm delineation errors are permissible for 8/15 BowelBag and Kidney_Left; 6/15 Kidney_Right and Liver; and 3/15 Stomach, SpinalCanal, and Duodenum cases.

Conclusion: The maximum permissible patient-specific OAR delineation error that ensures that the OAR dose remains below fDVT for pancreatic cancer cases was evaluated by using margin-expanded OARs. The maximum permissible delineation error varies with patient and OAR. Knowledge of the permissible delineation error enables reducing the OAR review workload in automated workflows in which automated contours are used with automated dose estimates prior to delineation review.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email