Radiobiologische effektbasierte optimierung von therapieplänen mit dem linear-quadratischen modell

Translated title of the contribution: Radiobiological effect based treatment plan optimization with the linear quadratic model

Stefan Schell, Jan J. Wilkens, Uwe Oelfke

Research output: Contribution to journalArticlepeer-review

8 Scopus citations


As an approach towards more biology-oriented treatment planning for external beam radiation therapy, we present the incorporation of local radiation damage models into three dimensional treatment planning. This allows effect based instead of dose based plan optimization which could potentially better match the biologically relevant tradeoff between target and normal tissues. In particular, our approach facilitates an effective comparison of different fractionation schemes. It is based on the linear quadratic model to describe the biological radiation effect. Effect based optimization was integrated into our inverse treatment planning software KonRad, and we demonstrate the resulting differences between conventional and biological treatment planning. Radiation damage can be analyzed both qualitatively and quantitatively in dependence of the fractionation scheme and tissue specific parameters in a three dimensional voxel based system. As an example the potential advantages as well as the associated risks of hypofractionation for prostate cancer are analyzed and visualized with the help of effective dose volume histograms. Our results suggest a very conservative view regarding alternative fractionation schemes since uncertainties in biological parameters are still too big to make reliable clinical predictions.

Translated title of the contributionRadiobiological effect based treatment plan optimization with the linear quadratic model
Original languageGerman
Pages (from-to)188-196
Number of pages9
JournalZeitschrift fur Medizinische Physik
Issue number3
StatePublished - Aug 2010
Externally publishedYes


Dive into the research topics of 'Radiobiological effect based treatment plan optimization with the linear quadratic model'. Together they form a unique fingerprint.

Cite this