Study and introduction of knowledge-based planning (KBP) models for VMAT treatment plans into clinical use

Postgraduate Thesis uoadl:3221418 9 Read counter

Unit:
Κατεύθυνση Ιατρική Φυσική-Ακτινοφυσική
Library of the School of Health Sciences
Deposit date:
2022-06-22
Year:
2022
Author:
Fanou Anna-Maria
Supervisors info:
Καλλιόπη Πλατώνη, Επίκουρη Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Ευστάθιος Ευσταθόπουλος, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Βασίλειος Κουλουλίας, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Original Title:
Study and introduction of knowledge-based planning (KBP) models for VMAT treatment plans into clinical use
Languages:
English
Translated title:
Study and introduction of knowledge-based planning (KBP) models for VMAT treatment plans into clinical use
Summary:
Purpose: This study focused on the creation of knowledge-based treatment planning (KBP) models utilizing a commercial system (RapidPlan) for head and neck cancer patients. Moreover, a comparison in terms of dosimetry and deliverability was performed between RapidPlan plans and the clinical ones.
Materials and Methods: Three RapidPlan head and neck models were created with 43 clinically acceptable Volumetric Modulated Arc Therapy (VMAT) plans (21 larynx, 11 nasopharynx and 11 oropharynx) as an initial training set for each model. Plans with different dose prescription were not mixed into the same model. The models were created and named based on the dose prescription to the Planning Target Volume (PTV), namely head and neck model 54Gy, 60Gy and 70Gy. The potential outliers were explored and the real ones were excluded from the models. The validation process was divided into closed loop with 10 cases (included in the training set) and open-loop with 21 cases (not included in the training set). The validation plans optimized with RapidPlan models were compared with the clinical ones in terms of dosimetry and plan deliverability.
Results: The training of the head and neck model 54 Gy demonstrated that 10 out of 15 trained Organs At Risk (OARs) structures achieved a coefficient of determination above 0.7. In head and neck model 60 Gy, 16 out of 18 trained OARs achieved a coefficient of determination above 0.7. In head and neck model 70 Gy, all trained OARs achieved a coefficient of determination above 0.7. The dosimetric assessment of the open loop validation set, showed that PTV coverage was comparable to the clinical and sparing in the 24 evaluated OARs was comparable or better compared to the clinical. Concerning Patient Specific Quality Assurance (PSQA), for 3%/3mm criterion, the 2D gamma passing rates for RapidPlan plans were higher from the clinical ones for phase 70 Gy (99% versus 98% - statistically significant) and had no statistically significant difference for phases 54Gy and 60Gy.
Conclusion: During validation it was found that RapidPlan plans were deliverable and dosimetrically comparable with the clinical ones for both the OARs and the PTV, but there is room for improvements.
Main subject category:
Health Sciences
Keywords:
RapidPlan, Knowledge-based planning (KBP), Head and neck cancer, VMAT
Index:
No
Number of index pages:
0
Contains images:
Yes
Number of references:
56
Number of pages:
73
File:
File access is restricted only to the intranet of UoA.

thesis-RapidPlan.pdf
3 MB
File access is restricted only to the intranet of UoA.