Database Plan Quality Impact on Knowledge-based Radiation Therapy Treatment Planning of Prostate Cancer
Master of Science (MS)
Physics and Astronomy
Purpose: Knowledge-based planning (KBP) leverages plan data from a database of previously treated patients to inform the plan design of a new patient. This work investigated bladder and rectum dose-volume prediction improvements in a common KBP method using a Pareto plan database in VMAT planning for prostate cancer. Methods: We formed an anonymized retrospective patient database of 124 VMAT plans for prostate cancer treated at our institution. From these patient data, two plan databases were compiled. The clinical plan database (CPD) contained planning data from each patient’s clinical plan, which were manually optimized by various planners. The multi-criteria optimization database (MCOD) contained Pareto plan data from plans created using a standardized MCO protocol. Overlap volume histograms, incorporating fractional OAR volumes only within the treatment fields, were computed for each patient and used to match new patient anatomy to similar database patients. For each database patient, CPD and MCOD KBP predictions were generated for D_10, D_30, D_50, D_65, and D_80 of the bladder and rectum in a leave-one-out manner. Prediction achievability was verified through a re-planning study on a subset of 31 randomly selected database patients using the lowest KBP predictions, regardless of plan database origin, as planning goals. Results: MCOD model predictions were significantly lower (p < 0.001) than CPD model predictions for all five bladder dose-volumes and rectum D_50 (p = 0.004) and D_65 (p < 0.001), while CPD model predictions for rectum D_10 (p = 0.005) and D_30 (p < 0.001) were significantly less than MCOD model predictions. KBP model predictions were statistically equivalent to re-planned values for all predicted dose-volumes, excluding D_10 of bladder (p = 0.03) and rectum (p = 0.04). Compared to clinical plans, re-plans showed significant average reductions in D_mean for bladder (7.8 Gy; p < 0.001) and rectum (9.4 Gy; p < 0.001), while maintaining statistically similar PTV, femoral head, and penile bulb dose. Conclusion: KBP dose-volume predictions derived from Pareto plans were lower overall than those resulting from manually optimized clinical plans. A re-planning study showed the KBP dose-volume predictions were achievable and led to significant reductions in bladder and rectum dose.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Wall, Phillip Douglas Hardenbergh, "Database Plan Quality Impact on Knowledge-based Radiation Therapy Treatment Planning of Prostate Cancer" (2017). LSU Master's Theses. 4443.