S1875
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
stable MTF (MD = 0.01 lp/mm), and improved contrast (MD = +4.16 lp/mm) on corrCBCTs. Conclusion: CorrCBCT-based dose recalculation using the RayStation workflow demonstrated robustness and accuracy for offline adaptive radiotherapy. Although minor HU variations were observed, their dosimetric impact remained negligible, supporting reliable dose evaluation across treatment sites. Optimization of imaging parameters (mA, rotation range, filtering) and careful FOV management remain essential to maintain consistent image quality. Further investigations are warranted to assess the impact of delineation accuracy on corrCBCT-based workflow. Keywords: Offline adaptative radiotherapy, corrected CBCT CBCT deformable registration to quantify dose variations for IMRT breast treatment plans Thahabah Mohammed Alharthi 1 , Awatef Hadi Aldawsari 2 , Ahmad Nashat Shamout 2 , Eyad Ali Alhakeem 2 1 Department of clinical laboratory sciences, College of Applied Medical Sciences, Taif University, Taif, Saudi Arabia. 2 Medical Physics Department, Taif Military Hospitals-Alhada, Taif, Saudi Arabia Purpose/Objective: To quantify the dosimetric effect of deformation for breast cancer patients during radiotherapy treatment course based on CBCT images when using IMRT Digital Poster 2148 technique. The changes in the dosimetric parameters and impact of skin flash (SF) for both chest wall (CW) and whole breast (WB) cases were evaluated. Material/Methods: A total of 23 breast cancer patients (including 33 CBCTs) were selected for retrospective study. IMRT plans for 17 CW and 6 WB were evaluated for dose variations during the course of treatment. Eclipse v16.1 registration and deformation algorithm [1] was used to generate deformable planning CT (dCTs). The deformation was conducted between CBCTs images and original planning CTs (pCTs) to propagate the PTVs and OAR contours. The deformed structures were reviewed by a radiation oncologist. The original IMRT plan was copied to the deformed (dCTs) set for each case and the dose was recalculated on the dCTs to extract dosimetric parameters. The effect of skin flash was also investigated. Results: Relative difference in the mean PTV D95% between the dCTs and pCTs calculations was -2.7% when using skin flash compared to -6.9% without skin flash. For OAR, the heart mean dose and ipsilateral lung V20 had no
Digital Poster 2130
Dosimetric and image quality assessment of corrected cone-beam CT for offline adaptive radiotherapy Maud Suszko 1 , Jimmy Fontaine 1 , Agathe Leroux 1 , Mariane Nourieh 2 , Julien Langrand-Escure 3 , Omar Jmour 3 , Jean-Baptiste Guy 3 1 Medical Physics Unit, Centre Marie Curie, Valence, France. 2 UFR PHITEM, Université Grenoble Alpes, Grenoble, France. 3 Radiation Oncology Unit, Centre Marie Curie, Valence, France Purpose/Objective: Corrected cone-beam CT (corrCBCT) represents a promising alternative to conventional re-scan CT (rCT) for treatment evaluation, enabling offline adaptive workflows on standard linacs. Based on an analytical image correction approach, this technique may facilitate dose recalculation and plan adaptation without the need for additional CT acquisitions. This study aimed to evaluate the Hounsfield Unit (HU) consistency, image quality, and dose calculation accuracy of corrCBCTs. Material/Methods: CorrCBCTs were evaluated on RayStation v23B. HU accuracy was assessed using a CIRS electron density phantom scanned on three Elekta linacs (Synergy and Versa HD) with thorax, pelvis, and head-and-neck imaging protocols. Planning CTs (pCTs) and corrCBCTs were compared using the Mean Absolute Error (MAE) for each insert. A HU–electron density (CT-to-ED) calibration curve was derived to quantify deviations and evaluate their dosimetric impact across 30 clinical VMAT plans (10 prostate, 10 breast, and 10 head-and- neck). A second cohort of 10 patients with same-day CBCT and rCT acquisitions was analyzed using dose– volume histogram (DVH) comparisons and 3D gamma analysis (3 %/3 mm and 2 %/2 mm). Image quality was further evaluated with a Catphan 600 phantom based on quantitative metrics—signal-to-noise ratio (SNR), modulation transfer function (MTF), and contrast—for both CBCT and corrCBCT datasets. Results: HU discrepancies on corrCBCTs remained within tolerance limits (±20 HU for values < 100 HU; ±50 HU for values > 100 HU), except for dense bone (maximum MAE = 202 ± 147 HU) and lung inserts (maximum MAE = 47.3 ± 62.1 HU). CT-to-ED calibration analysis indicated dose deviations below 1 % for all PTVs and OARs. For the second cohort, gamma passing rates ranged from 95–99 % (3 %/3 mm) and 80–97 % (2 %/2 mm). Values below 95 % were mainly related to pCT stitching and setup variations between CBCT and rCT; these findings were consistent with DVH analysis. Image quality assessment showed slightly lower SNR (mean difference, MD = –16.47 %),
Made with FlippingBook - Share PDF online