Gross tumor volume (GTV) delineation is central for radiotherapy planning. It provides the basis of the clinical target volume and, ultimately, the planning target volume which is used for dose optimization. Manual GTV delineations are prone to intra- and inter-observer variation and automatic segmentation methods also produce different results. There is no consensus on how to account for the contouring uncertainty, but has been suggested to incorporate it into the planning target volume (PTV) margin. Current recipes for the PTV margin are based on normal distribution assumptions and are more suitable for setup and execution errors. In this study we use the GTV delineations made by 6 experienced clinicians to create delineation-specific dose plans. These dose plans are then used to calculate theoretic tumor control probabilities (TCP) differences between delineations. The results show that current margin recipes are inadequate for maintaining the same TCP despite manual delineation variation. New methods to account for delineation variation should be developed. © Springer-Verlag 2013.
CITATION STYLE
Hollensen, C., Persson, G., Højgaard, L., Specht, L., & Larsen, R. (2013). Differences in radiotherapy delivery and outcome due to contouring variation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7761 LNCS, pp. 122–129). Springer Verlag. https://doi.org/10.1007/978-3-642-38079-2_16
Mendeley helps you to discover research relevant for your work.