Treatment-position magnetic resonance imaging reduces bony-landmark co-registration error in head and neck radiotherapy
Physics and Imaging in Radiation Oncology · 2026-11 · 42 · 101121
DOI: 10.1016/j.phro.2026.101121 ↗Kurz gefasst
Fachartikel in der Zeitschrift Physics and Imaging in Radiation Oncology, 2026-11. Verfasst von Tomohiro Itonaga und 8 weiteren Autor:innen. Die Arbeit nennt 16 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background and purpose: Routine diagnostic magnetic resonance imaging is acquired without immobilisation, creating co-registration uncertainty with planning computed tomography. We quantified whether treatment-position imaging reduced bony-landmark co-registration error, and whether the benefit persisted after resampling and deformable refinement. Materials and methods: Fifty patients with paired treatment-position and diagnostic magnetic resonance imaging were retrospectively analysed. Each magnetic resonance imaging dataset was separately registered to planning computed tomography. Mean Euclidean distance between corresponding bony landmarks on the computed tomography and magnetic resonance images was compared after rigid registration and after deformable refinement, using Holm-adjusted landmark-wise comparisons. Through-plane resampling was assessed in all 50 patients, and interobserver reproducibility in a 15-patient subset.
Results
Treatment-position imaging reduced mean Euclidean distance from 7.5 to 4.1 mm after rigid registration (difference 3.4 mm; 95% confidence interval 2.7–4.1; p = 2.0 × 10 −13 ). The benefit persisted after resampling (4.3 mm) and deformable refinement (3.4 versus 6.0 mm; difference 2.6 mm; 95% confidence interval 2.1–3.1; p = 7.1 × 10 −14 ), although refinement reduced the advantage by 0.8 mm (interaction p = 0.023). Intraclass correlation coefficients were 0.81 for diagnostic imaging and 0.48 for treatment-position imaging (95% confidence interval − 0.05–0.79), with small mean bias and lower treatment-position error in all 15 patients.
Conclusions
Treatment-position magnetic resonance imaging reduced bony-landmark error in co-registration with planning computed tomography after rigid registration, and the benefit persisted after deformable refinement. Because delineation and dose-volume parameters were not assessed, these findings should not be interpreted as demonstrating improved contouring accuracy.
Zitieren
APA
Itonaga, T., Mikami, R., Okubo, M., Kawaguchi, M., Zama, T., Kurooka, M., Araki, Y., Sugahara, S., & Saito, K. (2026). Treatment-position magnetic resonance imaging reduces bony-landmark co-registration error in head and neck radiotherapy. Physics and Imaging in Radiation Oncology, 42, 101121. https://doi.org/10.1016/j.phro.2026.101121
Vancouver
Itonaga T, Mikami R, Okubo M, Kawaguchi M, Zama T, Kurooka M, et al. Treatment-position magnetic resonance imaging reduces bony-landmark co-registration error in head and neck radiotherapy. Physics and Imaging in Radiation Oncology. 2026;42:101121. doi:10.1016/j.phro.2026.101121
BibTeX
@article{Itonaga2026,
author = {Itonaga, Tomohiro and Mikami, Ryuji and Okubo, Mitsuru and Kawaguchi, Masumi and Zama, Tatsuhiko and Kurooka, Masahiko and Araki, Yoichi and Sugahara, Shinji and Saito, Kazuhiro},
title = {Treatment-position magnetic resonance imaging reduces bony-landmark co-registration error in head and neck radiotherapy},
journal = {Physics and Imaging in Radiation Oncology},
year = {2026},
volume = {42},
pages = {101121},
doi = {10.1016/j.phro.2026.101121}
}
EUVIMED