Premium
Quantitative analysis of trabecular microstructure by 400 MHz nuclear magnetic resonance imaging
Author(s) -
Chung HsiaoWen,
Wehrli Felix W.,
Williams John L.,
Kugelmass Steven D.,
Wehrli Suzanne L.
Publication year - 1995
Publication title -
journal of bone and mineral research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.882
H-Index - 241
eISSN - 1523-4681
pISSN - 0884-0431
DOI - 10.1002/jbmr.5650100518
Subject(s) - materials science , magnetic resonance imaging , signal to noise ratio (imaging) , tomography , partial volume , nuclear magnetic resonance , microstructure , resolution (logic) , nuclear medicine , perimeter , image resolution , signal (programming language) , biomedical engineering , optics , physics , radiology , mathematics , geometry , medicine , artificial intelligence , computer science , metallurgy , programming language
A new approach for the quantitative analysis of trabecular microstructure, based on high‐field proton nuclear magnetic resonance (NMR) imaging, is presented. NMR is ideal because it provides high contrast between the marrow proton signal and the bone, which appears with background intensity. Images from 1 cm 3 defatted specimens of trabecular bone, suspended in water doped with 1 mM Gd(DTPA) to shorten T 1 to about 300 ms, can be obtained at a resolution on the order of 30–50 μm and slice thickness of 150 μm, in 10 minutes at 400 MHz proton frequency. Digital image processing algorithms were designed and evaluated for the measurement of bone area fraction, perimeter length, mean trabecular thickness, and separation. Bone area fraction derived from the NMR images was found to be in excellent agreement with bone volume fraction measured independently (slope = 0.96, r 2 = 0.924, p < 0.0001). Errors in the mean trabecular thickness and separation were <6%. The effects of finite imaging slice thickness and signal‐to‐noise ratio (SNR) were also evaluated. The data suggest a resolution of 50 × 50 × 200 μm 3 and an SNR on the order of 10 to provide safe margins for precise and accurate structural analysis by means of the algorithms presented in this paper. The method allows simultaneous measurement at multiple locations within the specimen volume without the need for physical sectioning.