Yan Y, Shu H, Bao X, Luo L, Bai Y
Department of Biomedical Engineering, Southeast University, Nanjing, China.
Int J Radiat Oncol Biol Phys. 1997 Aug 1;39(1):247-54. doi: 10.1016/s0360-3016(97)00305-2.
This article presents a new optimization method for stereotactic radiosurgery treatment planning for gamma unit treatment system.
The gamma unit has been utilized in stereotactic radiosurgery for about 30 years, but the usual procedure for a physician-physicist team to design a treatment plan is a trial-and-error approach. Isodose curves are viewed on two-dimensional computed tomography (CT) or magnetic resonance (MR) image planes, which is not only time consuming but also seldom achieves the optimal treatment plan, especially when the isocenter weights are regarded. We developed a treatment-planning system on a computer workstation in which Powell's optimization method is realized. The optimization process starts with the initial parameters (the number of isocenters as well as corresponding 3D isocenters' coordinates, collimator sizes, and weight factors) roughly determined by the physician-physicist team. The objective function can be changed to consider protection of sensitive tissues.
We use the plan parameters given by a well-trained physician-physicist team, or ones that the author give roughly as the initial parameters for the optimization procedure. Dosimetric results of optimization show a better high dose-volume conformation to the target volume compared to the doctor's plan.
This method converges quickly and is not sensitive to the initial parameters. It achieves an excellent conformation of the estimated isodose curves with the contours of the target volume. If the initial parameters are varied, there will be a little difference in parameters' configuration, but the dosimetric results proved almost to be the same.
本文提出一种用于伽马刀治疗系统立体定向放射外科治疗计划的新优化方法。
伽马刀已用于立体定向放射外科约30年,但医生-物理师团队设计治疗计划的常规程序是试错法。在二维计算机断层扫描(CT)或磁共振(MR)图像平面上查看等剂量曲线,这不仅耗时,而且很少能实现最佳治疗计划,尤其是在考虑等中心权重时。我们在计算机工作站上开发了一个治疗计划系统,其中实现了鲍威尔优化方法。优化过程从医生-物理师团队大致确定的初始参数(等中心数量以及相应的三维等中心坐标、准直器尺寸和权重因子)开始。目标函数可以更改以考虑对敏感组织的保护。
我们使用训练有素的医生-物理师团队给出的计划参数,或者作者大致给出的参数作为优化程序的初始参数。优化后的剂量学结果显示,与医生的计划相比,高剂量体积与靶体积的贴合度更好。
该方法收敛迅速,对初始参数不敏感。它实现了估计的等剂量曲线与靶体积轮廓的极佳贴合。如果初始参数发生变化,参数配置会有一点差异,但剂量学结果几乎证明是相同的。