Back to Search Start Over

A reusable 3D printed brain‐like phantom for benchmarking electrical properties tomography reconstructions.

Authors :
Meerbothe, T. G.
Florczak, S.
van den Berg, C. A. T.
Levato, R.
Mandija, S.
Source :
Magnetic Resonance in Medicine; Nov2024, Vol. 92 Issue 5, p2271-2279, 9p
Publication Year :
2024

Abstract

Purpose: In MR electrical properties tomography (MR‐EPT), electrical properties (EPs, conductivity and permittivity) are reconstructed from MR measurements. Phantom measurements are important to characterize the performance of MR‐EPT reconstruction methods, since they allow knowledge of reference EPs values. To assess reconstruction methods in a more realistic scenario, it is important to test the methods using phantoms with realistic shapes, internal structures, and dielectric properties. In this work, we present a 3D printing procedure for the creation of realistic brain‐like phantoms to benchmark MR‐EPT reconstructions. Methods: We created two brain‐like geometries with three different compartments using 3D printing. The first geometry was filled once, while the second geometry was filled three times with different saline‐gelatin solutions, resulting in a total of four phantoms with different EPs. The saline solutions were characterized using a probe. 3D MR‐EPT reconstructions were performed from MR measurements at 3T. The reconstructed conductivity values were compared to reference values of the saline‐gelatin solutions. The measured fields were also compared to simulated fields using the same phantom geometry and electrical properties. Results: The measured fields were consistent with simulated fields. Reconstructed conductivity values were consistent with the reference (probe) conductivity values. This indicated the suitability of such phantoms for benchmarking MR‐EPT reconstructions. Conclusion: We presented a new workflow to 3D print realistic brain‐like phantoms in an easy and affordable way. These phantoms are suitable to benchmark MR‐EPT reconstructions, but can also be used for benchmarking other quantitative MR methods. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
07403194
Volume :
92
Issue :
5
Database :
Complementary Index
Journal :
Magnetic Resonance in Medicine
Publication Type :
Academic Journal
Accession number :
179169153
Full Text :
https://doi.org/10.1002/mrm.30189