Program Information
A Simple Predictor for V105 in Breast Tangential Treatment Planning
H Chen, J Kim , D Carlson , J Deng , R Nath , Z Chen*, Department of Therapeutic Radiology, Yale Univ. School of Medicine, New Haven, CT
Presentations
SU-K-FS1-4 (Sunday, July 30, 2017) 4:00 PM - 6:00 PM Room: Four Seasons 1
Purpose: The purpose of this study is to demonstrate the feasibility of knowledge-based prediction of V105 for breast tangential treatment planning.
Methods: 45 previously treated breast conserving radiotherapy cases (20 right supine, 14 Left supine, and 11 prone) were used in this study (all plans used tangential fields and the field-in-field technique.) Energies used for these plans were 6MV, or 6MV mixed with 10MV or 15MV. PTV coverage was normalized to a 95% prescription dose that covers 95% of the PTV volume. The correlations between V105 and four patient-specific geometric parameters (PTV Volume (PV), irradiated Volume (IV), chest wall separation (CWS), and separation at half breast height (HHS)) were analyzed, respectively, to build a prediction model for V105. Suboptimal plans identified by the developed knowledge-based V105 prediction model were re-planned for model validation.
Results: Since breast volume and shape are different for each patient, the HHS may vary when the PV or IV remains same. We found that the V105 significantly correlated with the HHS separation(r =0.8, p <0.01). The Knowledge-based V105 prediction model based on this patient cohort being: “if HHS<17.5cm, then V105 ≤ 5% is achievable”. 3 suboptimal plans were identified, which the model predicted that hot spot of V105 could be lowered to less than 5%. Re-plans were performed and verified that V105 < 5% was achievable for the suboptimal plans, by using optimal energies and/or optimal MLC shapes.
Conclusion: This study demonstrated the knowledge-based prediction of V105 for breast tangential planning is feasible. A simple predicator for V105 serves as a quality control tool to identify the suboptimal plans. Subsequent re-planning demonstrated improved plan quality regarding the hot spot V105. The model can be further refined for other breast OAR dose matrix predictions, not only limited for V105.
Contact Email: