"accurate and precise delivery using robotically controlled linear accelerators and collimation systems; " imaging modalities capable of providing accurate volumetric anatomical and physiological data for treatment planning; " software systems that exploit imaging to construct virtual patient models; " numerical algorithms for simulating and optimizing dose distributions during planning; and " data and networking standards for the verify-andrecord software used to manage patient-specific information and to transmit complex treatment plans.

Exploiting the integration of these advances has led to the development of novel therapeutic approaches, such as intensity modulated radiation therapy (IMRT).

Advances in Treatment Planning
Computed tomography (CT) imaging has shifted the focus of treatment planning and guidance from inferring disease location based on radiographic bony landmarks to a more direct method of using softtissue to define both the tumour target and the normal organs in three dimensions. Magnetic resonance imaging (MRI), positron emission tomography (PET), and single positron emission computed tomography (SPECT) are being increasingly used to augment the process with more anatomical detail, and with physiological data. The importance of volumetric imaging in treatment planning cannot be over-emphasized. The radiographic approach is expedient, but requires large radiation fields to ensure that all of the diseased tissue is included in the treated volume. This is carried out at the cost of including excess normal tissue and limiting the radiation dose prescription to prevent unreasonable side effects. In contrast, the anatomically conformal approach attempts to maximally avoid normal tissues, reducing the toxic effects of treatment and enabling dose escalation.

After the appropriate imaging has been performed, treatment planning proceeds in the patient s absence. Imaging data are transferred to the planning computer, and are used to create a virtual patient model. In a largely manual process, physicians delineate target volumes and the essential normal tissues that may be affected by treatment toxicity.The direct delineation of anatomical structures in three dimensions supports a confined-field approach in which a multi-leaf collimator (MLC) shapes beams to conform to the target volume. The ability to precisely locate diseased tissue and delineate normal structures leads to a geometric approach in directing therapeutic beams, with multiple beams traversing the body from a wide range of angles and converging on the target. Target coverage can be optimized with beams directed to avoid sensitive normal tissues. It is equally important to note that the segmentation of images into anatomical structures also supports detailed dose accounting to accompany records of target and normal tissue effects monitored in clinical trials, and in the follow-up of individual patient outcomes.