S2081
Physics - Image acquisition and processing
ESTRO 2026
Purpose/Objective: Accurate quantification of CT numbers, expressed in Hounsfield Units (HU), is essential for diagnostic confidence, tissue characterization, and quantitative imaging applications, including radiotherapy dose calculation. This study aims to evaluate and compare the accuracy and reproducibility of CT numbers between Single-Energy CT (SECT), Dual-Energy CT (DECT) and Photon-Counting CT (PCCT), with emphasis on their quantitative potential for radiotherapy
treatment planning. Material/Methods:
An Advanced Electron Density Phantom (Sun Nuclear, Melbourne, FL, USA) containing tissue-equivalent inserts was scanned using SECT, DECT and PCCT systems (all Siemens Healthineers, Forchheim, Germany) with tube voltages of 120 kV, 100/140 kV and QP 140 kV, respectively. PCCT datasets were reconstructed into Virtual Monoenergetic Images (VMIs) across a range of energy levels (40–190 keV). Circular regions of interest (ROI Ø 2 cm) were placed at four phantom positions to measure CT numbers for solid water (Fig. 1a) and compared to manufacturer- provided reference values across reconstruction parameters (Br40, Qr40, Sd40, Rho, and VMI). CT number accuracy and reproducibility were assessed via mean deviation and standard deviation metrics. Results: For SECT imaging (Br40, Qr40, Sd40), mean CT numbers were highly consistent across all ROI position in the phantom, with mean variability below 3 HU. Comparable deviations were obtained for the electron density reconstruction Rho in DECT imaging. In contrast, PCCT demonstrated position- and energy- dependent variability in CT numbers (Fig. 1b). Across VMI levels, deviations ranged from 51 HU at 40 keV to 18 HU at 190 keV. Lower mean CT numbers were measured at peripheral phantom positions, particularly for VMIs levels between 40 and 60 keV. Minimal deviation was seen for VMI level 72 keV (Fig. 1c).
Conclusion: SECT demonstrated stable and reproduceable CT numbers across different phantom positions. Spectral CT imaging, especially PCCT, showed greater variability in CT number accuracy depending on ROI position and VMI level. Accurate HU calibration across energy levels is essential for precise dose calculation and robust quantitative imaging. Variations in CT numbers may cause uncertainties in dose calculation accuracy. Phantom-based validation of CT number accuracy is an essential step for the integration of spectral CT into routine workflows. Keywords: CT number accuracy, spectral CT, phantom study tomography images for nasopharyngeal carcinoma radiotherapy using a two-stage latent diffusion model Zhan Liu 1 , Mengyu Hao 2 , Zijie Mo 3 , Lecheng Jia 3 , Guanqun Zhou 2 1 College of Biomedical Engineering, Southern Medical University, Guangzhou, China. 2 Department of radiation Oncology, Sun Yat-sen University Cancer Center, Shanghai, China. 3 RT-CID, United Imaging Healthcare Co., Shanghai, China Purpose/Objective: This study aims to develop and evaluate a deep learning framework for predicting future fan-beam computed tomography (FBCT) images during radiotherapy for nasopharyngeal carcinoma (NPC). The specific objectives are: (1) Construct a two-stage latent diffusion model that integrates clinical textual Proffered Paper 3887 prediction of future fan-beam computed features and sequential FBCT scans to forecast anatomical changes; (2) Quantitatively evaluate the
Made with FlippingBook - Share PDF online