I233] Automated treatment planning and quality assurance

Abstract

In radiotherapy, there is a strong development striving at improving the quality of the treatment plans. Consequently, there is a great need for quality assurance procedures focusing on the treatment plan quality. The most fundamental aspect is perhaps the quality of a treatment plan for an individual patient (treatment plan optimization), which in turn is dependent on the quality of the underlying protocols and prioritization strategies (guidelines). However, there are also other important aspects, such as quality assurance of treatment planning algorithms and their application to specific diagnoses or treatment techniques (benchmarking), preparatory treatment planning for multi-centre clinical studies (dummy runs), and treatment planner achievements (training and competencies). The quality of a treatment plan may also encompass its deliverability at treatment (QC pass rate and plan robustness). To approach these issues in a quantitative fashion, there must be an unambiguous understanding of the concept of treatment plan quality, that can be applied in a continuous quality assurance program. While the plan quality in its fundamental sense is related to the degree of prescription fulfilment, it is also influenced by various limitations and uncertainties. In this respect, it is of foremost importance to use optimization objectives that are specific and measurable. Many of these issues can be addressed using automatically generated treatment plans, either data-base driven or protocol based, and extended to sets of multiple Pareto-optimal plans. In this presentation, we will review some of the possibilities for treatment plan quality assurance that becomes available with automated treatment planning techniques. The influence of short-cuts and limitations will be discussed.