临床应该使用global还是local gamma通过率？

最新推荐文章于 2025-05-04 22:46:35 发布

马上去吃汉堡王

最新推荐文章于 2025-05-04 22:46:35 发布

阅读量1.4k

点赞数 21

分类专栏：数学、物理与算法文章标签：算法

本文链接：https://blog.csdn.net/G751888629/article/details/136932335

版权

数学、物理与算法专栏收录该内容

12 篇文章

订阅专栏

预警：中英混杂知乎体。非常简略，详细内容可以直接去看给出的参考文献。

一、什么是gamma index

gamma index is used to Quality Assurance (QA). Agreement between measured and calculated dose is usually compared on a point-wise basis, evaluating both the absorbed dose and the location.

需要用到两个验证指标：Dose Difference (DD, unit %), Difference-to-agreement (DTA, unit mm)。

计算方法：Daniel A. Low 1997 “A technique for the quantitative evaluation of dose distributions” DOI: 10.1118/1.598248

简要介绍：IMRT Quality Assurance | Oncology Medical Physics

二、通过标准

TG119：

Tolerance limits: 3%/3mm, threshold 10%, gamma passing rate (GP%) >= 95%
Action limits: 3%/3mm, threshold 10%, GP% >= 88%

TG218：

Tolerance limits: 3%/2mm, threshold 10%, GP% >= 95%
Action limit: 3%/2mm, threshold 10%, %GP >= 90%

tolerance limits 比 action limits 更严格，如果临床计划没有通过 tolerance limits 也不代表这个计划在临床是不可接受的，还要看没通过的部位是否处于OARs或者靶区位置。

三、global & local dose normalization

The dose difference criterion is typically described as the percentage of the maximum dose for one or both of the dose distributions being compared (global normalization), or the percentage of the prescription dose. It also can be described as a local dose percentage (local normalization).

$DD_{global} = \frac{D_{calc}-D_{meas}}{D_{max}}$

$DD_{local} = \frac{D_{calc}-D_{meas}}{D_{calc}}$

很多论文都证实，选择使用global还是local会对%GP产生很大的影响。总的来说，使用local normalization会产生更低的%GP。例子如下：

For 2%/2 mm and 3%/3 mm (10% dose threshold), the prostate %GPs were 80.4% and 96.7% for global normalization and 66.3% and 90.8% for local normalization, respectively. On the other hand, the HN %GPs were 77.9% and 93.5% for global normalization and 50.5% and 70.6% for local normalization, respectively.

四、global or local ?

先说结论：临床上用global，机器调试或者故障排除的时候可以使用local。

TG218列出的原因：

Global normalization should be used. Global normalization is deemed more clinically relevant than local normalization. The global normalization point should be selected whenever possible in a low gradient region with a value that is ≥ 90% of the maximum dose in the plane of measurement. This will provide a more realistic measure of the comparison between the two dose distributions.

Local normalization is more stringent than global normalization for routine IMRT QA. It can be used during the IMRT commissioning process and for troubleshooting IMRT QA.

PS1: 补充一个有关%GP的插值问题。详细内容可查看TG-218的2.G.2 Spatial resolution这一节。

spatial resolution（空间分辨率，也可以说是体素大小voxel size）对计算%GP的准确性有很大的影响。如果不进行插值，gamma可能会在一些剂量梯度很陡的区域计算错误。

一个好的经验法则是评估剂量的分辨率不应大于 DTA 标准的 1/3。也就是说如果使用DTA = 3mm，那么体素的大小不应该大于1x1x1 mm3。如果体素的大小不够小，就需要使用插值来满足需求。

PS2: TG119和TG218都建议将剂量阈值设置为10%以去除没有临床相关性且会使分析产生偏差的低剂量区域。如果将这些低剂量区域包括在内，则会在使用global normalization的使用增加%GP。

2024.3.31