
==== Front
ArXiv
ArXiv
arxiv
ArXiv
2331-8422
Cornell University

arXiv:2409.05318v1
2409.05318
1
preprint
Article
Magnetization transfer explains most of the $T_1$ variability in the MRI literature
Assländer Jakob
9 9 2024
arXiv:2409.05318v1https://creativecommons.org/licenses/by/4.0/ This work is licensed under a Creative Commons Attribution 4.0 International License, which allows reusers to distribute, remix, adapt, and build upon the material in any medium or format, so long as attribution is given to the creator. The license allows for commercial use.
http://arxiv.org/abs/2409.05318v1
nihpp-2409.05318v1.pdf
Purpose: To identify the predominant source of the $T_1$ variability described in the literature, which ranges from 0.6 - 1.1 s for brain white matter at 3 T. Methods: 25 $T_1$-mapping methods from the literature were simulated with a mono-exponential and magnetization-transfer (MT) models, each followed by mono-exponential fitting. A single set of model parameters was assumed for the simulation of all methods, and these parameters were estimated by fitting the simulation-based to the corresponding literature $T_1$ values of white matter at 3 T. Results: Mono-exponential simulations suggest good inter-method reproducibility and fail to explain the highly variable $T_1$ estimates in the literature. In contrast, MT simulations suggest that a mono-exponential fit results in a variable $T_1$ and explain up to 62% of the literature's variability. Conclusion: The results suggest that a mono-exponential model does not adequately describe longitudinal relaxation in biological tissue. Therefore, $T_1$ in biological tissue should be considered only a semi-quantitative metric that is inherently contingent upon the imaging methodology; and comparisons between different $T_1$-mapping methods and the use of simplistic spin systems - such as doped-water phantoms - for validation should be viewed with caution.
==== Body
pmc
