Back to Search Start Over

Application of frozen Thiel-embalmed specimens for radiotherapy delineation guideline development: a method to create accurate MRI-enhanced CT datasets.

Authors :
Stouthandel MEJ
Pullens P
Bogaert S
Schoepen M
Vangestel C
Achten E
Veldeman L
Van Hoof T
Source :
Strahlentherapie und Onkologie : Organ der Deutschen Rontgengesellschaft ... [et al] [Strahlenther Onkol] 2022 Jun; Vol. 198 (6), pp. 582-592. Date of Electronic Publication: 2022 Apr 11.
Publication Year :
2022

Abstract

Purpose: Thiel embalming followed by freezing in the desired position and acquiring CT + MRI scans is expected to be the ideal approach to obtain accurate, enhanced CT data for delineation guideline development. The effect of Thiel embalming and freezing on MRI image quality is not known. This study evaluates the above-described process to obtain enhanced CT datasets, focusing on the integration of MRI data obtained from frozen, Thiel-embalmed specimens.<br />Methods: Three Thiel-embalmed specimens were frozen in prone crawl position and MRI scanning protocols were evaluated based on contrast detail and structural conformity between 3D renderings from corresponding structures, segmented on corresponding MRI and CT scans. The measurement error of the dataset registration procedure was also assessed.<br />Results: Scanning protocol T1 VIBE FS enabled swift differentiation of soft tissues based on contrast detail, even allowing a fully detailed segmentation of the brachial plexus. Structural conformity between the reconstructed structures on CT and MRI was excellent, with nerves and blood vessels imported into the CT scan never intersecting with the bones. The mean measurement error for the image registration procedure was consistently in the submillimeter range (range 0.77-0.94 mm).<br />Conclusion: Based on the excellent MRI image quality and the submillimeter error margin, the procedure of scanning frozen Thiel-embalmed specimens in the treatment position to obtain enhanced CT scans is recommended. The procedure can be used to support the postulation of delineation guidelines, or for training deep learning algorithms, considering automated segmentations.<br /> (© 2022. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany.)

Details

Language :
English
ISSN :
1439-099X
Volume :
198
Issue :
6
Database :
MEDLINE
Journal :
Strahlentherapie und Onkologie : Organ der Deutschen Rontgengesellschaft ... [et al]
Publication Type :
Academic Journal
Accession number :
35403891
Full Text :
https://doi.org/10.1007/s00066-022-01928-z