Knowledge-based Planning using Pseudo-structures for Volumetric Modulated arc Therapy (VMAT) of Postoperative Uterine Cervical Cancer: A Multi-Institutional Study

Abstract

Background: The aim of this study was to investigate the performance of the RapidPlan knowledge-based treatment planning system using models including registered pseudo-structures, and to determine how many structures are required for automatic optimization of volumetric modulated arc therapy (VMAT) for postoperative uterine cervical cancer. Methods: Pseudo-structures were retrospectively contoured for patients who had completed treatment at one of five institutions. For 22 patients, RPs were generated with a single optimization for models with two (RP_2), four (RP_4), or five (RP_5) registered structures, and the dosimetric parameters of these models were compared with a clinical plan with several optimizations. The total times for pseudo-structure creation and optimization were also measured.Results: Most dosimetric parameters showed no major differences between each RP. In particular, the rectum Dmax, V50Gy, and V40Gy with RP_2, RP_4, and RP_5 were not significantly different, and were lower than those of the clinical plan. In addition, the average proportions of plans achieving acceptable criteria for all dosimetric parameters were 98%, 99%, 98%, and 98% for the clinical plan, RP_2, RP_4, and RP_5, respectively. The average times for the creation and optimization of pseudo-structures were 105, 17, 21, and 29 minutes, for the clinical plan, RP_2, RP_4, and RP_5, respectively. Conclusions: The RapidPlan model with two registered pseudo-structures could generate clinically acceptable plans while saving time. This modeling approach using pseudo-structures could possibility be used for the VMAT planning process.

Multi-institution model (big model) versus single-institution model of knowledge-based volumetric modulated arc therapy (VMAT) planning for prostate cancer

“…Tol et al noted that the wide range of anatomical information in the RP model was important for generating better plan quality compared with the clinical plans 20 . The line objectives along the DVH lower bounds were also useful for optimizing the estimated DVHs predicted from the big model with the large number of combinations between anatomical and dosimetric characteristics of registered plans 15 , 21 , 22 . In the rectum, the big model could not reduce the differences in the V 90 and V 80 values between each institution compared with each single-institution model, as shown in Fig.…”