Exploratory evaluation of preserved human body-donor head specimens for quality assurance of synthetic computed tomography in cranial magnetic resonance-only radiotherapy

Axer BN, Jüngert K, Stache L, Paul M, Putz F, Jost T, Dickmann C, Brand J, Pieper L, Corradini S, Bräuer L, Paulsen F, Bert C, Szkitsak J (2026)

Publication Status: Submitted

Publication Type: Unpublished / Preprint

Future Publication Type: Journal article

Publication year: 2026

Original Authors: Bernd-Niklas Axer, Katharina Juengert, Lisa Stache, Miriam Paul, Florian Putz, Tina Jost, Christoph Dickmann, Johann Brand, Lucas Pieper, Stefanie Corradini, Lars Bräuer, Friedrich Paulsen, Christoph Bert, Juliane Szkitsak

DOI: 10.2139/ssrn.7370050

Abstract

Background and Purpose: In magnetic resonance (MR)-only radiotherapy, synthetic computed tomography (sCT) replaces computed tomography (CT) for dose calculation and patient positioning. Quality assurance (QA) surrogates with native bone morphology and complex tissue–air interfaces are lacking. Therefore, we evaluated preserved human body-donor head specimens for anatomy-realistic cranial sCT QA.Materials and Methods: Ten head specimens spanning four preservation pathways were imaged at 1.5 T: plastinated (n = 3), formalin-fixed and ethanol-preserved (n = 5), nitrite-pickling-salt-ethanol-polyethylene-glycol (NEP)-embalmed (n = 1), and fresh-frozen (n = 1). One plastinate was ferromagnetic and excluded. Six specimens were scanned three times, without and with re-setup. Two-dimensional (2D) slice-wise and three-dimensional (3D) volume-based algorithms received identical T1-weighted Dixon input. Bone and soft-tissue agreement was quantified by Dice similarity coefficient (DSC), 95th-percentile Hausdorff distance (HD95), Hounsfield unit (HU) mean error (ME) in the sCT-derived and reference-CT-derived masks, and dose-volume histogram deviations.Results: Bone DSC was highest for a formalin-fixed head (0.72) and lowest for a plastinate (0.11). Bone ME was positive under sCT-derived and negative under reference-CT-derived masking throughout. Mean dose deviation was smallest for a dissection-course specimen (1.3%); the fresh-frozen head failed dosimetrically (75%). Re-setup changed metrics more than repetition alone. 3D improved all bone metrics (p ≤ 0.039) but not soft-tissue or dose agreement.Conclusions: Preserved head specimens can complement phantoms for anatomy-realistic cranial sCT QA. Suitability appears to depend on complete neurocranial anatomy, fixation preserving MR signal, absence of ferromagnetic components, room-temperature scanning, and standardized positioning.

Authors with CRIS profile

Bernd-Niklas Axer Department of Radiation Oncology Katharina Jüngert Lehrstuhl für Funktionelle und Klinische Anatomie Lisa Stache Lehrstuhl für Funktionelle und Klinische Anatomie Miriam Paul Lehrstuhl für Funktionelle und Klinische Anatomie Florian Putz Department of Radiation Oncology Tina Jost Department of Radiation Oncology Johann Brand Department of Radiation Oncology Lucas Pieper Department of Radiation Oncology Stefanie Corradini Department of Radiation Oncology Friedrich Paulsen Lehrstuhl für Funktionelle und Klinische Anatomie Christoph Bert Department of Radiation Oncology Juliane Szkitsak Department of Radiation Oncology

Involved external institutions

Siemens Healthineers (Germany) DE Germany (DE)

How to cite

APA:

Axer, B.-N., Jüngert, K., Stache, L., Paul, M., Putz, F., Jost, T.,... Szkitsak, J. (2026). Exploratory evaluation of preserved human body-donor head specimens for quality assurance of synthetic computed tomography in cranial magnetic resonance-only radiotherapy. (Unpublished, Submitted).

MLA:

Axer, Bernd-Niklas, et al. Exploratory evaluation of preserved human body-donor head specimens for quality assurance of synthetic computed tomography in cranial magnetic resonance-only radiotherapy. Unpublished, Submitted. 2026.

BibTeX: Download