S2040
Physics - Image acquisition and processing
ESTRO 2026
References: 1 Wang et al. "TrackRAD2025 challenge dataset: real- time tumor tracking for MRI-guided radiotherapy." (2025) Medical Physics. DOI: 10.1002/mp.17964 Keywords: MRI-linac,data compression,segmentation Digital Poster 100 Radiotherapy CT simulation. HU-to- relativeElectronDensity calibration curves. A validation study Periklis M Papavasileiou 1,2 , Dimitris Markellos 2 , Anna Makridou 2,3 , Thaleia Karamarkou 2 , Michael Chatzimarkou 2,3 1 Biomedical Sciences, University of West attica, Athens, Greece. 2 Department of Radiotherapy and Radiosurgery, St Luke's Hospital, Thessaloniki, Greece. 3 Medical Physics, Theageneio Oncology Hospital, Thessaloniki, Greece Purpose/Objective: In radiotherapy, CT number to relative electron density (EDr) curves are required for dose distribution calculation. The aim of this study is to validate the CT- 2-EDr calibration curves generated using two different phantoms. Material/Methods: The CATPHAN 604 and the CIRS AEDP phantoms, with ten (EDr range 0.001-1.868) and sixteen (EDr range 0.28-1.78) tissue inserts respectively, were scanned on the Siemens-Go-Sim scanner using abdominal CT protocol (120kVp, 145mAs eff, Safire 3 & QR4 reconstruction kernel, reconstructed slices with 2mm3 isotropic voxels). Following reconstruction, circular ROIs were drawn on the different structures generating VOIs of 1cm3. The average CT value of each VOI was computed and then plotted against the corresponding EDr to generate the CT-2-EDr curves. For the CATPHAN 604 phantom, two linear fittings were carried out, one using all the ten structures and one using the eight structures that were common on the CATPHAN 500 series. For the AEDP phantom, two fittings were attempted. In the first fitting, a linear equation was fitted to all sixteen structures. In the second one, the structures were divided into two groups according to their EDr, 0.28- 1.05 and 1.16-1.78 respectively, and a linear fit was employed for each group. Then, the CT-2-EDr curves from each phantom were used to estimate the EDr values of the other and the % difference between the estimates and the true values was computed. Results: For the CATPHAN phantom, with the exception of the ‘Air’ insert (difference above 200% since air is lying at the extremity of the CT-2-EDr curve), the absolute difference between the AEDP-based and true EDr
values is below 17%. For tissues with EDr above 1.10, the EDr estimates using the bilinear fitting are more accurate, whereas for those below 1.10, the single fitting is more accurate.For the AEDP phantom, the absolute difference between the CATPHAN-based EDr estimates and true values is below 31%. For tissues with EDr above 0.5, the estimates based on fitting using all the CATPHAN structures are closer to the true values, compared to fitting using only eight structures. Similarly, for issues with EDr below 0.5, the fitting is more accurate when only the eight CATPHAN structures is employed. Conclusion: The results show that the EDr estimates, as computed by the CT-2-EDr calibration curves, are affected by the phantom employed and the attempted fitting. Further phantom work is required to study the effect on the generated dose distributions. Keywords: HU, calibration, CATPHAN Digital Poster Highlight 132 Superior Tissue Conspicuity in Photon-Counting CT Compared to Energy-Integrating Detector CT: Implications for Radiotherapy Contouring Wing Sze Wong, Tin Lok Chiu, Siu Ki Yu Medical Physics Department, Hong Kong Sanatorium & Hospital, Hong Kong, Hong Kong
Purpose/Objective: Accurate target delineation is essential for
radiotherapy planning. While contrast-enhanced CT improves visualization, contouring uncertainty persists due to limited soft tissue contrast in conventional energy-integrating detector CT (EID-CT). With direct photon-counting detectors, photon-counting CT (PCCT) has the potential to improve the visualization of both enhanced and non-enhanced tissues. This study investigates whether PCCT improves the inter-tissue contrast-to-noise ratio (CNR) compared to EID-CT in both phantom and preliminary clinical studies. Material/Methods: Iodine solutions (0-8 mg/mL) were prepared in plug and placed in a cheese phantom. The phantom was scanned on a PCCT (Siemens NAEOTOM Alpha) and an EID-CT (Siemens Somatom Drive) at 120 kVp and 140 kVp, with matched CTDIvol, pitch, and kernel. PCCT data were reconstructed as virtual monoenergetic images (VMI) at 40, 50, 60, and 70 keV, plus a conventional polyenergetic image (T3D). Plug pairs were used to simulate clinical contrasts (e.g., 3 vs. 0-2 mg/mL, 5 vs. 0-4 mg/mL, 8 vs. 0-5 mg/mL), mimicking tumors with high iodine uptake against surrounding normal tissues with lower uptake. Relative CNR differences between PCCT and EID-CT were quantified across energies and voltages. In the preliminary
Made with FlippingBook - Share PDF online