Microscopy in Magnetic Resonance Imaging
Introduction
Magnetic resonance imaging (MRI) is finding increasing application in areas that require microscopic resolution. While typical resolutions employed clinically are on the order of a millimeter, the notion of using MRI at microscopic resolutions arose early in the development of this technique.1, 2 The limitations to spatial resolution in MRI have been reviewed by several authors.3, 4, 5 Theoretically, resolution is limited by molecular diffusion and is estimated to be of order 1–10 μm.4 In practice, resolution is limited by the image signal-to-noise ratio (SNR). The achievable SNR is most often limited by the available time to acquire the image. For example, a clinical MR image with reasonable SNR can be obtained in about 5 min with a voxel (volume element) size of 1 mm3. If one were to acquire the same image with a 100 μm3 voxel size at the same SNR and all other experimental parameters constant (assuming this is possible), it would take approximately 10 years to complete the scan. The challenge in magnetic resonance microscopy (MRM) is to overcome the problem of excessive imaging time as the resolution is improved. This is achieved by optimizing the experimental setup, both hardware and software, to overcome the intrinsically poor SNR in order to obtain a respectable image in a reasonable amount of time. A number of laboratories now routinely obtain three-dimensional MR images of various samples with good SNR and contrast at spatial resolutions of 30–100 μm3.6, 7, 8, 9, 10, 11 Applications range over a wide spectrum from the geological12 to the biological.13, 14
For the purposes of this review we define ‘microscopic’ MRI as studies with spatial resolution on the order of 100 μm or less and will concentrate on developments since 1999. We recommend texts by Callaghan,5 Mansfield and Morris,15 Morris,16 Vlaardingerbroek and den Boer17 and Haacke et al.18 Two publications12, 19 contain papers presented at two conferences. Reviews by Balaban,20 Bhakoo,21 Blackband,6 and Glover and Mansfield22 cover earlier material. In the following sections we outline the basic challenge of improving signal to noise and the role of various factors that affect contrast and resolution in MRM images. Emphasis is laid on applications and illustrated with recent results. Our personal bias towards biological imaging is apparent and we refer the reader to texts by Blumich23 and Blumler12 for in-depth discussions of application in material sciences. We conclude with an evaluation of the prospects for routinely achieving higher resolution in MRM images.
Section snippets
Signal and noise in MRM
MRM places heavy demands on the signal detection system on account of the small magnitude of the initially generated signal that then undergoes further processing as demanded by the experimental protocol. In order to generate the initial signal, one depends on the magnetization from a sample containing nuclear spins (usually protons) placed in a strong magnetic field. The magnitude of this magnetization at thermal equilibrium24 is proportional to the number of spins in the sample as well as the
Contrast
Perhaps the most basic sample characteristic contributing to contrast is the variation in proton density across the sample. Voxel intensity is directly proportional to proton concentration, all other factors being, thus proton concentration differences between voxels give rise to the well-known spin density contrast. Magnetization transfer (MT) contrast is finding increasing application in the clinical realm, but thus far is little used in MRM.57
Imaging methods and pulse sequences
The MR characteristics of the sample to be imaged largely determine the suitability of an imaging procedure. The time required for imaging and the desired resolution need to be considered together. By reducing the demand on resolution, the number of points to be sampled in k-space can be reduced and hence the imaging time minimized. Our aim here is to highlight some of the points to be kept in mind in evaluating a method of imaging for a particular goal. Many of the pulse sequences chosen here
q-Space and displacement imaging
There is an implicit assumption of a single T2 and thus single D in Eq. (5), i.e., monoexponential behavior of signal decay in the diffusion experiment. If this is not true, q-space analysis is an efficient and useful method of data interpretation. This technique takes advantage of the Fourier relationship between the signal decay and the displacement probability function, , which gives the average probability of a particle having a dynamic displacement over a time t. The echo
Cellular imaging with MRM using T2 and contrast: magnetic labeling
T2 and contrast arising from the introduction of small paramagnetic particles inside a cell can facilitate visualization and serve as a useful marker in MRM. The magnetic particles give rise to a local field inhomogeneity and cause a reduction in the T2 of water protons undergoing translational diffusion near these particles. Spin echo (T2) and gradient echo based MR imaging sequences can be used to visualize the presence of small quantities of the particles, with gradient echo being
Prospects for Higher Resolution
The availability of higher field MRI magnets, improved receiver coil and gradient coil design and hardware are some of the factors that have led, in recent years, to an increase in sensitivity. This has naturally led to attempts by several investigators to obtain MRM images with increased resolution in reasonable times especially with a view to extend MRM to biology and in particular to image single cells. The earliest successful attempt in this direction was made by Aguayo et al.248 who
Acknowledgements
This work was funded in part by the National Institute of Biomedical Imaging and Bioengineering, the National Institute on Drug Abuse, the National Center for Research Resources, and the Beckman Institute.
References (292)
- et al.
Magma
(1999) - et al.
Biophys. J.
(1999) - et al.
Prog. Neurobiol.
(2002) - et al.
Neuroanatomical micromagnetic resonance imaging
- et al.
Neuroimage
(2003) - et al.
Prog. NMR Spectrosc.
(2002) - et al.
The application of magnetic resonance imaging and spectroscopy to gene therapy
- et al.
J. Magn. Reson.
(1984) - et al.
J. Magn. Reson.
(1989) - et al.
Magn. Reson. Imaging
(1992)
J. Magn. Reson. Ser. A
J. Magn. Reson. Ser. B
J. Magn. Reson.
J. Magn. Reson.
J. Magn. Reson.
J. Magn. Reson. Ser. B
J. Magn. Reson. Ser. B
J. Magn. Reson.
Magn. Reson. Imaging
J. Magn. Reson.
J. Magn. Reson.
Magn. Reson. Imaging
J. Magn. Reson.
J. Magn. Reson.
J. Magn. Reson.
J. Magn. Reson.
J. Magn. Reson. A
Magn. Reson. Imaging
Magn. Reson. Imaging
J. Magn. Reson.
J. Magn. Reson.
Magn. Reson. Imaging
Nature
Phys. Rev. B
IEEE Trans. Nucl. Sci.
Med. Phys.
Principles of Nuclear Magnetic Resonance Microscopy
Radiology
Spatially Resolved Magnetic Resonance: Methods, Materials, Medicine, Biology, Rheology, Geology, Ecology, Hardware
J. Agric. Food Chem.
NMR Imaging in Biomedicine
Nuclear Magnetic Resonance Imaging in Biology and Medicine
Magnetic Resonance Imaging: Theory and Practice
Magnetic Resonance Imaging: Physical Principles and Sequence Design
Magnetic Resonance Microscopy
ILAR J.
Rep. Prog. Phys.
NMR Imaging of Materials
The Principles of Nuclear Magnetism
J. Comput. Assist. Tomogr.
Cited by (9)
High-resolution NMR spectra in inhomogeneous fields utilizing the CRAZED sequence without coherence selection gradients
2008, Journal of Magnetic ResonanceDouble-quantum-filtered intermolecular single-quantum coherences in nuclear magnetic resonance spectroscopy and imaging
2006, Chemical Physics LettersCitation Excerpt :Intermolecular multiple-quantum coherences (iMQCs) [1–3], resulting from residual long-range dipolar interactions [4,5] in highly polarized nuclear spin systems such as water, have found important applications in magnetic resonance imaging (MRI) [6–8] and high-resolution nuclear magnetic resonance (NMR) [9–12].
Magnetic resonance imaging of water ascent in embolized xylem vessels of grapevine stem segments
2013, Canadian Journal of Plant ScienceMRI in developmental biology and the construction of developmental Atlases
2011, Cold Spring Harbor Protocols