Purpose: Precision manual contouring is a time- and labor-intensive process for radiotherapy that is subject to inter- and intra-observer variabilities. Furthermore, contours drawn prior to treatment do not necessarily represent the real patient anatomy over an extended treatment course. This study investigates the impact of auto-generated organ-at-risk (OAR) contours on treatment outcome.
Methods: Prostate, seminal vesicles (SV), bladder, and rectum were manually delineated on planning CT and daily CBCT images for 20 prostate patients. OARs were automatically generated with atlas-based segmentation algorithms on planning CTs. Treatment plans were created via a novel automated planning application accessing a knowledge-based planning solution for both auto-generated OARs (aOAR-plan) and manual OARs (mOAR-plan). Both low-risk patients (LRP, CTV=Prostate) and intermediate-risk patients (IRP, CTV=Prostate+SV) were studied. Calculated doses were transferred from planning CT to daily CBCTs based on clinical shifts, then contour-based deformable registrations were used to obtain the cumulative doses. Differences in cumulative dose for the bladder and rectum were used as an indicator of the agreement between auto-plans and manual-plans. Furthermore, 5 patients were selected to test the difference between obtaining cumulative dose from transferred doses versus doses recalculated on CBCT. Maximum dose, V55, V60, V65, V70 and V75 were used for the dosimetric evaluation.
Results: For both LRP and IRP groups, good dosimetric agreement between the aOAR-plans and mOAR-plans was observed. The dose differences were within 2.5% for bladder and within 4% for rectum, respectively. No significant dose differences between the transferred and recalculated doses were observed for both the bladder and rectum.
Conclusion: Manually drawn precision contours are used to represent patient anatomy that often fluctuate by multiple millimeters on a daily basis, thereby making the time and effort questionable. Automated contouring tools offer improvement in contour consistency, and provide acceptable delivered doses compared with precision contours.
Funding Support, Disclosures, and Conflict of Interest: This research was supported by a grant from Varian Medical Systems, Palo Alto, CA.
Not Applicable / None Entered.
Not Applicable / None Entered.