Complexity score-based plan quality control of VMAT
10.3760/cma.j.cn113030-20210901-00338
- VernacularTitle:基于复杂性评分VMAT计划质量控制
- Author:
Jinyan HU
1
;
Liyuan ZHANG
;
Yangguang MA
;
Bin HAN
;
Yuexin GUO
Author Information
1. 郑州大学第一附属医院放射治疗部，郑州 450052
- Keywords:
Volumetric-modulated arc therapy;
Complexity metric;
Principal component analysis;
Complexity score
- From:
Chinese Journal of Radiation Oncology
2022;31(9):817-822
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To explore the difference in the complexity of different treatment planning systems, multi-leaf collimator (MLC) types and treatment sites of volume-modulated arc therapy (VMAT), and propose a complexity score for plan quality control.Methods:Statistical analysis of 12 complexity metrics including Monaco and Eclipse, Agility, Millennium and High-definition MLC, nasopharyngeal, lung and cervical cancer was performed. Spearman correlation coefficient between complexity metrics was calculated. Principal component analysis was conducted to reduce the dimensionality of the original data set to the first two principal components and explain its physical meaning. Complexity score based on the principal components was calculated to establish warning and action thresholds for plan quality control. The correlation between complexity metrics and γ pass rate was analyzed.Results:Except cervical cancer aperture sub-regions metric, other metrics had significant differences between Monaco and Eclipse. Monaco MLC had a more regular field but higher MU, smaller leaf gap, and longer leaf travel distance. High-definition MLC with smaller leaf width significantly added MLC aperture-related metrics. The first two principal components explained over 80% of the total variance of the original dataset, complexity score was weighted average of first two principal components. The distribution of complexity score for different equipment and sites was different. The warning threshold was expressed as the average plus standard deviation, and the action threshold was expressed as the average plus 2 standard deviations. Complexity metrics and complexity scores had small correlation with γ pass rate, showing weak or irrelevant but statistically significant. Conclusions:Different planning systems, MLC types, and treatment site complexity metrics are significantly different. The complexity score is a useful tool for plan quality control.
