5F04-Physics of Intensity Modulated Radiation Therapy (IMRT)
Lei Xing and Arthur Boyer
Department of Radiation Oncology, Stanford University
Stanford, CA 94305-5847, USA
Developments in computer plan optimization and computerized delivery of radiotherapy using multileaf collimators (MLC) have led to a new treatment modality referred to as intensity-modulated radiation therapy (IMRT). The advent of IMRT represents one of the most importance technical advancement in the history of radiation therapy. Institutions worldwide are attempting or planning to integrate this new technology into their clinics. The purpose of this talk is to present an overview of physics aspects of the state-of-the-art IMRT technology.
IMRT consists the following major steps: patient immobilization and treatment planning CT/MRI image acquisition, tumor target volume and sensitive structure delineation, inverse treatment planning, pre-treatment quality assurance (QA) and simulation on, on-treatment QA, and IMRT dose delivery. Among them, the treatment planning, QA and dose delivery are distinctly different from the conventional 3D conformal radiation therapy (CRT). Instead of using X-ray beams with either uniform or wedged-shaped photon fluence distributions as in 3D CRT, the fluence profile of an IMRT beam can take an arbitrary form and is optimally chosen on a patient specific basis. For a given patient, the task of IMRT treatment planning is to determine the fluence profiles of the incident beams (typically 5-9 beams from different directions). This is usually achieved by using a special computer optimization technique named inverse treatment planning, during which the computer iteratively computes and compares different competing IMRT plans based on a mathematical model that mimics the decision-making process of a physician in ranking different competing treatment plans. This process continues until a satisfactory plan is obtained. Several commercial companies now offer inverse treatment planning systems for IMRT and much research efforts are being made in both academic institutions and industries to improve the performance of the currently available computing techniques.
Upon the completion of dose optimization, each optimal fluence profile is converted into a stack of segmented MLC-shaped fields through a leaf sequencing software implemented in the treatment planning system. The MLC sequences specified by their shapes and relative weights are then compiled into an electronic file and the execution of the file at the treatment delivery machine will realize the intended intensity modulation. Intensity modulation is achieved with computer-controlled MLC using either static, in which the MLC movement from a segment to another and the dose delivery are done at different instances, or dynamic delivery techniques, in which the MLC movement and dose delivery are realized simultaneously.
A rigorous but practical QA procedure is critically important to ensure that the planned dose distribution can be achieved safely and accurately. The QA of IMRT has three natural aspects; commissioning and testing of the system, verification of treatment plans, and routine QA of the MLC delivery system. Principles and practice of the QA for radiotherapy can be found in the AAPM Task Group Report 40. The tasks of patient specific QA can be divided into geometric verification and dosimetric verification. The former is concerned with the geometric accuracy of IMRT beams, typically including the isocenter verification and the portal verification. A pair of orthogonal digital reconstruction radiographs (DRRs) are used to verify the patient position by comparison with simulation films from a conventional simulator. A “conventional” portal image for an IMRT field can be created as well, using the MLC boundary as the port of the radiation field. AAPM IMRT Subcommittee (Medical Physics 30, 2089-2115, 2004) and ASTRO IMRT Scope Committee (International Journal of Radiation Oncology, Biology, Physics 58, 1616-1634, 2004) have made a series of recommendations regarding what dosimetric quantities need to be examined in order to dosimetrically validate an IMRT treatment plan.
5F04-调强放疗物理方面的问题
Lei Xing and Arthur Boyer
Department of Radiation Oncology, Stanford University
Stanford, CA 94305-5847, USA
调强放疗（IMRT）这种新治疗技术的出现是由于计算机治疗计划优化和计算机控制的使用多叶准直器（MLC）实施放疗的发展，它是放疗历史上最重大的技术进步之一，全世界的放疗部门都在尝试或计划将这一新技术应用于临床。本讲的主要内容是讨论IMRT技术物理方面的问题。
IMRT包括以下主要步骤：患者固定与治疗计划CT/MRI图像的获取、肿瘤靶区和敏感组织结构的勾画、逆向治疗计划、治疗前的质量保证（QA）与模拟、治疗中的QA、及IMRT剂量实施。其中，治疗计划、QA和剂量实施与常规的3D适形放疗（CRT）有很大不同。3D CRT射野的X线光子注量分布是均匀的或楔形的，而IMRT的注量分布是不均匀的，是根据患者情况得到的最佳的注量分布。对一特定的患者，IMRT治疗计划的任务是要确定各个射野的注量分布（通常设置5—9个不同方向的射野），完成这一任务需要采用特殊的计算机优化技术——逆向治疗计划，根据一定的数学模型（数学模型模拟医生选择治疗计划的决策过程），计算机迭代计算并比较各IMRT计划，直到得到满意的计划。目前已有几家公司提供IMRT逆向治疗计划系统，学术机构和商家还在进行大量的研究工作，以提高计算技术的性能。
完成剂量优化后，通过运行治疗计划系统中的生成叶片序列软件，将各优化的注量分布转化成多个MLC子野，MLC序列的子野形状、权重编辑成电子文件，在治疗机上执行，实现所需要的强度调制。通过计算机控制MLC来实现静态或动态调强，静态调强时，MLC运动形成下一个子野，在MLC运动过程中不出束，子野形成后进行剂量实施；动态调强时，MLC运动和剂量实施同时进行。
严格且实用的QA是确保所计划的剂量分布得以安全、准确实现的关键。IMRT QA有3个方面的内容：系统的验收和测试、治疗计划验证及MLC实施系统的常规QA。放疗QA的原则和实践见AAPM TG40号报告。患者QA分为位置验证和剂量验证。位置验证验证IMRT射野几何位置的准确性，包括等中心验证和射野验证。一对正交的DRR图像与常规模拟机拍摄的模拟片进行比较来验证患者的摆位。可以使用MLC的边界作为常规射野，以得到IMRT射野的“常规”射野影像。AAPM IMRT分委会(Medical Physics 30, 2089-2115, 2004)和ASTRO IMRT委员会(International Journal of Radiation Oncology, Biology, Physics 58, 1616-1634, 2004)就IMRT治疗计划剂量验证的问题提出了一系列的建议。