Beam angle selection and rotational therapy
Beam angle selection
The irradiation angles of a treatment beam ensemble have a critical influence on the quality of a radiation therapy treatment plan. In clinical routine, however, patients are usually irradiated with equi-spaced beam configurations (black beam configuration in figure 1.). Rarely, the beam configuration is manually adjusted by a radiation oncologist in a tedious trial and error process. For a human expert, however, it is impractical to fully oversee the non-intuitive interdependence of beam angles and the corresponding photon fluence profiles due to the sheer complexity of the optimization problem under consideration.
Figure 1: Conventional equi-spaced coplanar beam orientations (black) and optimized beam orientations (red) for the treatment of a complex intra-cranial tumor (red).
© dkfz.de
In our group, we are investigating if and how it is possible to improve the quality of radiation therapy treatment plans using an automated optimization of beam angles within the computer-aided treatment planning process. We are investigating both
- fast heuristics based on geometric or dosimetric considerations to find beneficial beam ensembles prior to conventional fluence optimization [1] and
- global optimization techniques to consider the interdependence of beam orientaions and corresponding photon fluence profiles
Rotational therapy
The basic idea of rotational therapy (rIMRT) is not only to deliver individual field shapes at each gantry angle within a continuous gantry rotation but also to vary the applied dose rate during the dose delivery process. Exploring this additional degree of freedom it is possible to deliver highly conformal dose distribution usually within one or two gantry revolutions [2,3]. Typical delivery times are in the range of two to five minutes which is often faster than a conventional IMRT treatment with similar plan quality.
Conventional IMRT treatment planning systems do not take into account the additional parameters like variable gantry speed or dose rate during the optimization process. Therefore new optimization algorithms have been developed within our group and up to now successfully applied for the analysis of phantom irradiations using our research accelerator [4].
Since not all installed linear accelerator are capable of dynamically changing the leaf positions while rotating the gantry our group in collaboration with Siemens is investigating an approach which delivers the individual fields in a stroboscopic way (i.e. the radiation is only on when the leaves are not moving during the continuous rotation from one gantry angle to the next). We are not only developing quality assurance approaches for this treatment technique but also performing treatment planning and dosimetric studies comparing this approach to the other existing rIMRT methods.
Direct aperture optimization for rotational therapy
Conventional IMRT treatment planning involves two independent steps: an optimization of the fluence profiles followed by a sequenzer that translates the fluence profile into deliverable multileaf collimator (MLC) segments. The sequencing step may lead to a degradation of the plan quality, which has not been considered during the preceding optimization of the fluence profiles.
Direct aperture optimization (DAO) avoids the sequencing step by simultaneously optimizing the aperture weights and shapes (i.e. the leaf positions of the MLC). We are currently investigating DAO based on a gradient algorithm. We are especially interested in the application of DAO for rotational therapy, where many constraints on the MLC motion need to be considered during the optimization process.
People involved
- Mark Bangert, Postdoc
- Claas Wessels, MSc. student
Selected publications
[1] M. Bangert & U. Oelfke: Spherical cluster analysis for beam angle optimization in intensity-modulated radiation therapy treatment planning. Phys. Med. Biol., 55 (2010).
[2] S. Ulrich S, S. Nill & U. Oelfke Development of an optimization concept for arc-modulated cone beam therapy. Phys. Med. Biol., 52(14) (2007) 4099-4119.
[3] K. Otto Volumetric modulated arc therapy: IMRT in a single gantry arc. Med. Phys., Jan 35(1) (2008).
[4] S. Ulrich S Optimization, Realization and quality assessment of arc-modulated cone beam therapy. Dissertation Naturwissenschaftl.-Mathematische Gesamtfakultät, Universität Heidelberg, (2009).