ESTRO 2025 - Abstract Book

S4377

RTT - Treatment planning, OAR and target definitions

ESTRO 2025

3379

Proffered Paper Clinical feasibility of using a commercial synthetic-CT solution for brain MRI-Only radiotherapy treatment planning Lamyaa Aljaafari 1,2,3 , Richard Speight 2 , David L Buckley 1 , David Bird 2 1 Leeds Institute of Cardiovascular & Metabolic Medicine (LICAMM), University of Leeds, Leeds, United Kingdom. 2 Department of Medical Physics and Engineering, Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom. 3 Department of Diagnostic Radiology, King Saud bin Abdulaziz University for Health Sciences, Alahssa, Saudi Arabia Purpose/Objective: An MRI-only workflow relies on synthetic computed tomography (sCT) images for dose calculation. This study evaluates the clinical feasibility of a commercial artificial intelligence-based sCT solution by comparing the dosimetric accuracy of clinical plans on sCT and CT. Material/Methods: 93 retrospective patients with brain cancer treated with volumetric modulated arc therapy (VMAT) were included. The patients underwent both CT and MRI scans as part of their clinical pathway. sCT images were generated from the MRI using a commercial AI-based algorithm (magnetic resonance for calculating attenuation, MRCAT Brain, Philips) The sCT images were rigidly co-registered to the CT images and clinical contours transposed from the CT onto the sCT within RayStation. The clinical plan produced on the CT was recalculated on the sCT. Dosimetric accuracy was validated by assessing dose differences at clinically relevant points of dose volume histograms (DVH) for the planning target volume (PTV) and organ at risk (OAR). The 95% confidence intervals for the mean dose difference between the dose plan in CT and sCT were computed. Results: Figure 1 shows examples of axial slices in CT and sCT for three brain patients. The mean dose differences between the DVH statistics are shown in Table 1. The PTV mean dose differences between CT and sCT were 0.3%, 0.4%, and 0.2% for D50%, D2%, and D98%, respectively. Two patients have shown a dose difference of ±2.0% due to missing bone in the skull base and ethmoid bone. The OAR mean dose differences were less than 0.3% for all OARs. Notably, extreme outliers were observed in two patients who had dose differences larger than -2.0%. One patient showed dose differences of -3.6% and -7.6% in the right orbit and lens, respectively . This patient had an artifact in the eye region, which caused signal loss on the MRI and dosimetric errors on the sCT. Figure 2 illustrates the distribution of dose differences between sCT minus CT in DVH metrics, highlighting the observed variations.