Histological Validation of Multi‐Echo Gradient Echo (MGRE)‐Derived Myelin Water Fraction (MWF) at 9.4 T and the Influence of Orientation on Quantification

• Published in: NMR in biomedicine Vol. 38; no. 1; pp. e5303 - n/a
• Main Authors: Thompson, Garth John, Wang, Ziyi, Kim, Jae‐Yoon, Li, Hui, Kim, Dong‐Hyun, Ye, Qiong, Su, Min‐Ying
• Format: Journal Article
• Language: English
• Published: England 01.01.2025
• Subjects: Adult; Aged; Animals; Body Water - diagnostic imaging; Female; Humans; Magnetic Resonance Imaging; Male; Middle Aged; Myelin Sheath - chemistry; Myelin Sheath - metabolism; Rats; Rats, Sprague-Dawley; Reproducibility of Results; Water - chemistry
• ISSN: 0952-3480; 1099-1492; 1099-1492
• DOI: 10.1002/nbm.5303

ABSTRACT Myelin is essential in the nervous system of mammals. As the location and degree of myelin loss can reflect varied pathophysiological status, noninvasive measurement of myelin is of high importance. The magnetic resonance imaging (MRI) technique of myelin water fraction (MWF) derived from multi‐echo gradient echo (MGRE) sequence is a promising tool for the quantification of myelin content due to the low specific absorption rate (SAR) compared with the spin‐echo sequence, time efficiency, and wide availability. Yet to our knowledge, MGRE‐derived MWF has never been quantitatively validated with histology. The main objective of this study was to quantitatively validate the MRI findings by referencing the myelin histology using a rat model. As a second objective, we investigated how the orientation of white matter fibers with respect to the static B0 field impacted both the apparent transverse relaxation rate (R2* = 1/T2*) and the derived MWF. Moreover, MWF is known to change with age; thus, we compared rat brains of different ages. The orientation effect of MWF in a clinical setting was studied using 3 T human data. Twenty ex vivo rat brains with different ages and three healthy volunteers were scanned on a 9.4 T Bruker and 3.0 T Siemens systems, respectively. The 3D MGRE and diffusion tensor imaging (DTI) data were acquired. Our results showed a highly significant correlation between MGRE‐derived MWF and histological stain of myelin, and susceptibility and diffusivity also demonstrated a significant association with myelin. Both MWF and R2* (R2* = 1/T2*) values changed as a function of orientation, and the function varied with age. Furthermore, MWF and R2* were more sensitive to age than DTI. In vivo 3 T human MWF also changed substantially with the orientation as well. Our results support that MGRE‐derived MWF can be used to assess the myelin content quantitatively. Coregistered histological LFB stain of myelin and MR parametric maps of the same slice of rat 8M1. (A) The original LFB stain, (B) optical attenuation of LFB stain, which reflects the concentration of myelin, (C) MGRE‐derived MWF using rNNLS algorithm (∈01$$ \in \left[0\ 1\right] $$, unitless), (D) FA map (∈01$$ \in \left[0\ 1\right] $$, unitless), (E) QSM map (ppm$$ \mathrm{ppm} $$), (F) RD map (×10−3mm2/s$$ \times {10}^{-3}{\mathrm{mm}}^2/\mathrm{s} $$). Thirty‐three manually outlined ROIs were projected to all these coregistered maps. The placement of ROIs is identical in (B) and (C).