Search

Background and Purpose: MRI-based synthetic CTs (synCTs) show promise to replace planning CT scans in various anatomical regions. However, the head-and-neck region remains challenging because of patient-specific air, bone and soft tissues interfaces and oropharynx cavities. Zero-Echo-Time (ZTE) MRI can be fast and silent, accurately discriminate bone and air, and could potentially lead to high dose calculation accuracy, but is relatively unexplored for the head-and-neck region. Here, we prospectively evaluated the dosimetric accuracy of a novel, fast ZTE sequence for synCT generation., Materials and Methods: The method was developed based on 127 patients and validated in an independent test (n = 17). synCTs were generated using a multi-task 2D U-net from ZTE MRIs (scanning time: 2:33 min (normal scan) or 56 s (accelerated scan)). Clinical treatment plans were recalculated on the synCT. The Hounsfield Units (HU) and dose-volume-histogram parameters were compared between the synCT and CT. Subsequently, synthetic treatment plans were generated to systematically assess dosimetry accuracy in different anatomical regions using dose-volume-histogram parameters., Results: The mean absolute error between the synCT and CT was 94 ± 11 HU inside the patient contour. For the clinical plans, 98.8 % of PTV metrics deviated less than 2 % between synCT and CT and all OAR metrics deviated less than 1 Gy. The synthetic plans showed larger dose differences depending on the location of the PTV., Conclusions: Excellent dose agreement was found based on clinical plans between the CT and a ZTE-MR-based synCT in the head-and-neck region. Synthetic plans are an important addition to clinical plans to evaluate the dosimetric accuracy of synCT scans., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2025. Published by Elsevier B.V.)