Indirect detection of biological tissue microstructure by nuclear magnetic resonance microscopy
Abstract
The microstructure of biological tissues is closely associated with their underlying physiological function. Therefore, tissue morphology has proved useful in disease classification. Nuclear magnetic resonance microscopy (NMRM), defined as magnetic resonance imaging at microscopic resolution (1μm to 0.1mm) offers excellent tissue contrast. However, advances in high resolution in NMRM have been hampered by the intrinsically low detection sensitivity (expressed in terms of signal-to-noise (SNR)). To overcome these problems, researchers have resorted to specialized hardware, including high-field superconducting magnets and radiofrequency microcoils. The current technology allows for imaging voxel sizes on the order of 100 μm3 at the scan time of an hour. ^ The goal of this dissertation is to explore NMR imaging techniques, where microstructural information can be retrieved without resolving the underlying structures, thereby greatly mitigating the SNR problem. Specifically, two imaging methods were investigated: q-space and intermolecular multiple quantum coherence (iMQC) imaging. NMR q -space imaging utilizes restricted diffusion to probe quasi-regular structures. Diffraction patterns observed from the signal attenuation reveal the characteristic size of structural elements. Q-space imaging was applied to differentiate white matter fiber tracts of rat spinal cord that differ in their axonal architecture. The experimental results were compared with those from simulations on optical microscopic images using an existing finite-difference diffusion model. The data demonstrate that calculated mean displacements and kurtosis parallel mean axon size and axonal density. Distant dipolar couplings between pairs of spins separated by a correlation distance lead to the observation of iMQC signals. Structural regularity then causes a signal modulation from which the spatial period of the sample can be derived. The method has been evaluated by means of a phantom of periodic spin density. Finally, the potential of this method to assess trabecular bone architecture was explored, where results suggesting regional mean trabecular plate separation can be estimated. ^
Subject Area
Engineering, Biomedical|Health Sciences, Radiology|Biophysics, Medical
Recommended Citation
Chih-Liang Chin,
"Indirect detection of biological tissue microstructure by nuclear magnetic resonance microscopy"
(January 1, 2004).
Dissertations available from ProQuest.
Paper AAI3125800.
http://repository.upenn.edu/dissertations/AAI3125800