S2116
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
To quantitatively validate our metric, we created synthetic inter-fraction datasets for 16 oesophageal and 5 pancreatic cancer patients (all prescribed 40Gy). For this, the image of the first fraction was registered using a contour-guided algorithm to images of 4 subsequent fractions, yielding registered images with known displacements resembling the actual clinical inter-fraction displacements. The known displacements (and their inverses) were used to create delineations, doses, and benchmark accumulated doses for each registered fraction image.Doses from each simulated dataset were accumulated using each fraction as a reference, employing three different deformable registration algorithms [1-3]. The benchmark dose accumulation error (DAE) was calculated as the mean pairwise difference between the DVH-parameters from the benchmark and estimated accumulated doses for each fraction. Results: The figure below shows the performance of our DACE for the oesophageal cancer patients. Each point denotes the DACE and benchmark DAE value of one patient for one of the DVH-parameters and one of the registration algorithms. A strong relationship (r=0.92) is seen for all DVH-parameters and registration algorithms. The mean absolute difference between the DACE and benchmark DAE is a factor of 3.0 lower than the difference with the conventional scenario of assuming a dose accumulation error of zero.
For pancreatic cancer patients, we found r=0.64, and factor 2.3 lower error. Interestingly, this improvement holds when using only 3 fractions (factors 2.1 and 1.5,
respectively). Conclusion:
We present a quantitative, automatic, and patient- specific metric for quality assurance of inter-fraction deformable dose accumulation. It is agnostic to the registration algorithm, anatomy, imaging modality (with consistent delineations available) and requires no training or parameter tuning. It strongly predicts the benchmark dose accumulation error. References: 1. Horn BK, Schunck BG. Determining optical flow. Artificial Intelligence. 1981;17:185-203.2. Thirion JP. Image matching as a diffusion process: an analogy with Maxwell’s demons. Medical Image analysis. 1998;2(3):243-60.3. Klein S, Staring M, Murphy K, Viergever MA, Pluim JP. Elastix: a toolbox for intensity- based Medical image registration. IEEE Transactions on Medical Imaging. 2009;29:196-205. Keywords: dose accumulation, evalution, quality assurance
Made with FlippingBook - Share PDF online