ESTRO meets Asia 2024 - Abstract Book

S302

Physics – Detectors, dose measurement and phantoms

ESTRO meets Asia 2024

183

Digital Poster

The effect of auto-planning parameter settings on the quality of automated cervical cancer plans Mengze Li

Department of Radiation Oncology, Chongqing University Cancer Hospital, Chongqing, China

Purpose/Objective:

This study aims to refine the parameters of automatic planning to improve plan quality and enhance the

protection of organs at risk (OARs).

Material/Methods:

A total of 30 cases were included for the design of both manual and automated volumetric modulated arc therapy (VMAT) plans. Two categories of parameters were involved: 2 qualitative parameters and 3 quantitative parameters. The qualitative parameters consist of the Engine Type, a control optimization engine with only two options, “Biological” (default) and “Non-Biological”, and the Use Cold-Spot ROIs, which adjusts target volume uniformity with choices “Yes” (default) and “No”. The quantitative parameters include Tuning Balance, which balances target volume and OARs and is adjustable in the range of [1%, 100%] with a step size of 10%, having a default value of 11%. A higher value shifts optimization focus gradually from target volume to OARs. The Dose Fall-off Margin, adjusting target conformity with a range of [1.0cm, 4.5cm] and a step size of 0.5cm, has a default value of 2.6cm. The Hot-Spot Maximum Goal, controlling the hot spot, has a range of [101%, 110%] with a step size of 1% and a default value of 107%. The original automated plan with default parameters, the automated plan with adjusted parameters, and the manual plan were compared for differences in the target conformity index, homogeneity index, and dosimetric parameters for OARs. This comparison aimed to determine the optimal parameter settings. Additionally, a comparative analysis of planning complexity was conducted among the manual plan, the original automated plan, and the adjusted automated plan. Apart from Engine Type, adjustments to the other four parameters will impact the Conformity Index (CI), Homogeneity Index (HI), and maximum dose (Dmax) for the target volume. Engine Type will affect the average dose to OARs. Hot-Spot Maximum will influence the Dmax for OARs overlapping with the target volume, such as the bowel. Compared to the manual plan, the original automated plan exhibited some shortcomings in protecting OARs. However, when adjusting parameters to "No" for Use Cold-Spot ROIs, setting Tuning Balance to 100%, and adjusting Hot-Spot Maximum Goal to 110%, while keeping the remaining parameters at their default settings, the automated plan demonstrated more optimal protection of OARs. In comparison to the original automated plan, the adjusted automated plan showed only a slight increase in Dmax for the bowel (approximately 2.11%, P<0.05). On the other hand, various OARs exhibited varying degrees of dose reduction. Notably, there were statistically significant reductions (P<0.05) in the V35, V40 of rectum, the V25, V30, V35of femoral heads, and the V10, V15, V20 of kidney, all exceeding 30% reduction. Results:

Conclusion: