The present invention relates to methods, systems, software arrangements and storage medium for measuring a magnetic field correlation function in a magnetic resonance imaging area, and more particularly, to methods for measuring the magnetic field correlation (“MFC”) function utilizing asymmetric spin echoes.
A theory of using asymmetric spin echoes is described, in part, in Jensen J. H. et al. Chandra R. Magn Reson Med; 44:144 (2000) (the “Jensen Publication”), the entire disclosure of which is incorporated herein by reference.
The local magnetic field experienced by a water proton is quantitatively characterized by the magnetic field correlation function, K(t), which may be defined by
K(|t−t′|)=B(t)B(t′)
[1]
with B(t) being the difference, at a time t, between the magnitude of the local field and the magnitude of the spatially uniform main field, B0, and with the angle brackets indicating an averaging over all the water protons within a given region of interest. K(t) may depend on both the spatial distribution of the magnetic field inhomogeneities and the diffusional dynamics of water molecules. The magnetic field correlation function provides information beyond the information contained in the standard nuclear magnetic resonance (“NMR”) relaxation times. Thus, the present invention provides a novel technique to, inter alia, examine the properties of tissue, blood, iron-rich regions of the brain, and tumors.
Although it is believed that the MFC has not been specifically measured, it was introduced at least as early as 1953 in the seminal work of Anderson and Weiss on NMR line shapes. See Anderson P W, Weiss P R, Rev Mod Phys 1953; 125:269-276 (the “Anderson Publication”), the entire disclosure of which is incorporated herein by reference. More recently, it has been utilized in several studies on the modeling of MRI contrast. For example, see Callaghan P T, Oxford University Press, New York, 1991; (the “Callaghan Publication”) Kennan et al., Magn Reson Med 1994; 131:9-21 (the “Kennan Publication”); Stables et al., Magn Reson Med 1998; 140:432-442 (the “Stables Publication”); and the Jensen Publication et al., the entire disclosures of which are incorporated herein by reference.
The MFC may be described as the magnetic resonance (“MR”) signal intensity as a function of the acquisition time, and can be approximated by the exponential form:
K(t)=K0 exp(−t/τ), [2]
where K0=K(0) is the magnetic field variance and i is a characteristic decay time. It has recently been shown in the Jensen Publication, that when water diffusion is only weakly restricted the MFC is more accurately described by an algebraic expression of the form:
There is not a significant amount of quantitative information regarding the MFC, except for an exact result that can be derived for an idealized random sphere model.
Asymmetric single spin echoes were first introduced by Dixon and Sepponen. In particular, Dixon demonstrated how asymmetric spin echoes can be used to separate the MR signals originating from water and fat. See Dixon W T, Radiology 1984; 1153:189-194 (the “Dixon Publication”), the entire disclosure of which is incorporated herein by reference. Sepponen et al., applied asymmetric spin echoes to obtain chemical shift images. See Sepponen R E, et al., Comput. Assist. Tomography 1984; 18:585-587 (the “Sepponen Publication”), the entire disclosure of which is incorporated herein by reference. The asymmetry of the Dixon sequence arises by shifting the signal acquisition time, while the asymmetry of the sequence of Sepponen et al. is achieved by shifting the 180° refocusing pulse.
Asymmetric single spin echo technique with a shifted refocusing pulse has been previously used to obtain MFC data. See Wismer et al., J Comput Assist Tomography 1988; 112:259-265 (the “Wismer Publication”); Rosenthal et al., Invest Radiology 1990; 1025:173-178 (the “Rosenthal Publication”); Thulborn et al., Am J. Neuroradiol. 1990 (the “Thulborn Publication”); 11:291-297; Hoppel et al., Magn Reson Med 1993; 130:715-723 (the “Hoppel Publication”); Ganesan et al., J. Magn. Reson. (B) 1993; 1102:293-298 (the “Ganesan Publication”); and the Stables Publication, the entire disclosures of which are incorporated herein by reference. However, there remains a need for improved methods of measuring the magnetic field correlation.
The present invention relates to methods, systems, software arrangements and storage medium for measuring the magnetic field correlation function, and more particularly, to methods, systems, software arrangements and storage medium measuring the magnetic field correlation function utilizing asymmetric spin echoes.
In an exemplary embodiment of the present invention, method, systems, software arrangements and storage medium obtaining the magnetic field correlation (“MFC”) of a sample using magnetic resonance for imaging (“MRI”) are provided that apply multiple spin echo sequences in which at least one spin echo sequence is an asymmetric spin echo sequence. Resultant information may then be acquired and the MFC can be determined as a function of the acquired information.
In a further exemplary embodiment of the present invention, the spin echo sequences may be an Asymmetric Dual Spin Echo Sequence (“ADSE”) having multiple echoes. In another exemplary embodiment of the present invention, the spin echo sequences may be an Echo Planar Imaging-Asymmetric Dual Spin Echo Sequence (“EPI-ADSE”) having multiple echoes.
In a further embodiment of the present invention, the asymmetric spin echo sequence can be applied by shifting a refocusing pulse such that the time between the rotation pulse and the refocusing pulse (t1) is not equal to the time between the refocusing pulse and the signal acquisition (t2), e.g., t1≠t2. In yet another exemplary embodiment, the asymmetric spin echo sequence may be applied by shifting the acquisition of the resultant information such that the time between the rotation pulse and the refocusing pulse (t1) is not equal to the time between the refocusing pulse and the signal acquisition (t2), i.e., t1≠t2.
In a further exemplary embodiment, the MFC can be determined as a function of the resultant information by applying the formula
in which γ is the proton gyromagnetic ratio, Sn is the signal intensity of the nth echo; and ts=|t1−t2| with t1 being the time between a rotation pulse and a refocusing pulse, and t2 being the time between the refocusing pulse and a signal acquisition.
a) shows an exemplary illustration of randomly distributed spheres representing cells with a higher magnetic susceptibility then the surrounding tissue;
b) shows exemplary contour lines for the magnetic field shifts induced by an external field oriented in the z direction, with solid lines provided for positive shifts, and dotted lines provided for negative shifts;
c) shows diffusion paths for two water molecules beginning at a time t and ending at a time t′;
d) shows an illustration of an exemplary magnetic field correlation (“MFC”) as a function of time representing a decay of the field shift correlation;
a) shows an exemplary asymmetric spin echo signal intensity provided as a function of the refocusing pulse time shift for two regions of interest within a subacute hemorragic brain lesion; and
b) shows an exemplary semi-logarithmic graph provided as a function of the square of the time shift
Other and further objects, features and advantages of the present invention will be readily apparent to those skilled in the art upon a reading of the description of preferred embodiments which follows.
The present invention relates to methods, systems, software arrangements and storage medium for measuring the MFC function, and more particularly, to methods, software arrangements and storage medium for measuring the MFC function utilizing asymmetric spin echoes.
Since the MFC is sensitive to the spatial distribution of microscopic magnetic field inhomogeneities, MFC imaging enables another way to probe tissue microstructures. For example, MFC imaging may be potentially useful for evaluating tissues with substantial intrinsic microscopic magnetic susceptibility variations. It may therefore be used as a sensitive probe of tissue microstructure that provides certain information beyond what is contained in the standard nuclear magnetic resonance (“NMR”) relaxation times, such as T2 and T2*. In addition, microscopic susceptibility variations can be created by the introduction of a contrast agent. For example, the contrast agent may be a paramagnetic element, e.g., gadopentetate dimeglumine (Gd-DTPA). Therefore, the imaging of the MFC may be useful in the evaluation of neurodegenerative disorders, such as Parkinson's and Alzheimer's diseases, since ferritin (i.e., the most prevalent form of iron in the brain) is paramagnetic and localized in oligodendrocytes, which have diameters of about 7 μm. Indeed, it is known that the T2 relaxation rate, as measured with a Carr-Purrcell-Meiboom-Gill (CPMG) sequence, depends significantly on the interecho time in the brain region. Thus, the MFC with a nontrivial time dependence may be expected and the measurement of this time dependent MFC can provide a significant marker of disease state.
The MFC imaging may be of particular interest for Alzheimer's disease, as conventional relaxation time measurements may not be as useful. Another possibly important application the MFC imaging may be to the imaging of tumors. Most paramagnetic contrast agents, such as Gd-DTPA, enter only the extra cellular spaces of tissues, and thereby create microscopic magnetic susceptibility variations. The MFC of a contrast agent infused tumor may be sensitive to tumor morphology (e.g., cell size and density). The measurement of the MFC can therefore assist in tumor identification and grading. Moreover, measuring the MFC with the magnetic resonance imaging (“MRI”) may be a preferable way for probing inhomogeneities on a length scale of a few tens of micrometers. This may be because MRI echo times are typically in the range of 10 ms to 100 ms, and the water diffusion constant in biological tissues is about 1.0 μm2/ms. This implies that the temporal behavior of the measured MFC may be sensitive to spatial variations in the range of 8 μm to 25 μm. Biologically, this may be an important length scale, since this is the size of capillaries and many types of cells.
The MFC may be defined as the product of the magnetic field shifts experienced by a water molecule at two different times, averaged over all the water molecules within a specified region of interest. The field shifts may be the differences between the magnitude of the field and the magnitude of the uniform background field (which may be nearly the same as the external field magnitude). If the biological tissue has linear magnetic characteristics, as is usually the case, then the MFC may vary as the square of the external field. The dependence of the MFC on the difference in the two sampling times, on the other hand, may be purely an intrinsic tissue property. In practice, the MFC may be most sensitive to the spatial dependence of the field shifts over length scales of about 10 μm, which is comparable to the size of many cell types.
To appreciate the significance of the MFC, it is preferable to understand the definition of the MFC, which is provided as follows. For example, a tissue may be exposed to a strong, uniform external magnetic field, as is the case in a typical MRI experiment. Because of spatial variations in its magnetic susceptibility, the tissue may generate an inhomogeneous secondary field. The susceptibility variations can be provided due to structures such as iron-rich cells or vessels with a deoxygenated blood. A water molecule diffusing through the tissue may experience a field that varies with time due to field inhomogeneities. If δB(t) is defined as the difference at time t between the magnitude of the field at the water molecule and the uniform background field, then the MFC may be defined by
K(|t−t′|)=δB(t)δB(t′
[4]
where the angle brackets indicate an average over all the water molecules in a specified region of interest (in practice this may be an MRI voxel). The temporal behavior of the MFC, which is indicated with a symbol K in mathematical expressions, involves the time difference |t−t′| as follows from an assumption that the state of the tissue may be time invariant. It should be noted that, by definition, the MFC can depend quadratically on the magnitude of the field inhomogeneities.
As an illustration,
According to one exemplary embodiment according to the present invention, the MFC at specific times can be directly imaged without recourse to any particular model. This exemplary technique thereby yield information beyond what may be provided by conventional relaxation time measurements. The fact that the MFC contains information distinct from T2 and T2* can be described using the following simple example. For example, two test tubes, A and B, can be provided each containing an aqueous suspension of microspheres and equal concentrations of a paramagnetic contrast agent. If the microspheres in test tube A have a radius of 10 μm and the microspheres in test tube B have a radius of 20 μm, the spatial density of spheres in test tube A can be eight times higher than in test tube B. The volume fraction filled by the spheres can then be approximately the same in both suspensions. T2 can be the same for both test tubes because T2 can be determined primarily by the contrast agent concentration. T2* can also be essentially the same, because T2* may depend primarily on the contrast agent concentration and the volume fraction of the spheres. Therefore, the test tubes may appear approximately identically on T2 or T2* weighted images. However, if a water diffusion constant of 1.0 μm2/ms is assumed, then for a time difference of 20 ms the MFC values for the two test tubes will differ by about a factor of two, and the test tubes may appear sharply different on an MFC image.
A relationship between the MFC and transverse relaxation rates may be most clearly revealed through a connection between the MFC and T2 as measured with a Carr-Purcell-Meiboom-Gill (CPMG) sequence. Because of water diffusion through field gradients, the measured T2, as obtained from a monoexponential fit to the echo intensities, may depend on the interecho time, 2Δt. In the limit 2Δt→0, the true T2 value may be obtained. If the magnetic field inhomogeneities are not too large, then one can show that T2 as a function of Δt may be provided by
with γ representing the proton gyromagnetic ratio.
If the full dependence of T2 on the interecho time were available, then it is possible to invert Eq. [5] to determine the MFC. However, it is generally only practical to obtain T2 for a limited number of Δt values, and so a straightforward inversion may not be possible. In some cases, it may be possible to postulate a model form for the MFC, having a few adjustable parameters, and then fit the model to the T2 data by using equation [5]. However, the success of this approach may depend on the accuracy of the model, which can be difficult to establish.
The physical significance of the MFC can be twofold. First, the magnitude of the MFC may provide a measure of the amplitude of the magnetic field inhomogeneities within a tissue. In particular, as follows from Eq. [3], K(0) can be the variance of the field. Second, the decay of the MFC with time may contain information about the spatial variation of the field inhomogeneities. In particular, the temporal rate of change of the MFC at a time t (i.e., K′(t)) can be primarily sensitive to field inhomogeneities with a length scale of about √{square root over (6Dt)}, where D is the water diffusion constant, where √{square root over (6Dt)} being the average distance a water molecule diffuses over a time t. Thus, the MFC may be measured in conjunction with diffusion-weighted imaging (DWI) in order to obtain an estimate of the water diffusion constant. This additional information may then allow for a more accurate assessment of the length scale associated with the field inhomogeneities. If the inhomogeneities vary only on length scales that are large compared to about √{square root over (6Dt)} (i.e., >>25 μm), the MFC can be essentially independent of time and may be related to the NMR line width. Particularly, the second moment of the NMR spectral line may be approximately about 4γ2K.
According to an exemplary embodiment of the present invention, the MFC may be determined by applying 2 or more spin echo sequences, in which at least one spin echo sequence is an asymmetric spin echo sequence.
where γ is the proton gyromagnetic ratio.
Equation [6] is based on a weak field approximation, but is also usable at clinical field levels of a few Tesla or less. The correlation function enters equation [6] preferably only for times between t−ts and t+ts. If K(t) is approximated in this temporal range by a linear function, the following formula may be applied:
Hence, the ratio of the signal intensities for a standard spin echo sequence and for an asymmetric spin echo sequence can be used to estimate K(t) at a selected time. This exemplary technique may be utilized for multiple spin echo sequences so that K(t) at several different times may be determined from one pair of acquisitions.
For example, the key may be a relationship between the MFC and the signal intensity for an asymmetric CPMG sequence. The asymmetric CPMG sequence differs from a conventional CPMG sequence in that the 180° refocusing pulses are shifted from their usual times by an amount ts, as is illustrated in
with the convention S−1(ts)=1. The corrections to equation [8] are preferably third order in the field, and for weak fields are small compared to the MFC, which is second order in the field. The integral in equation [8] is over the time t ranging from −ts to ts. As long as ts is not too large, it is reasonable to make the following linear approximation:
K[(t+2n−1)Δt]≈K[(2n−1)Δt]+tK′[(2n−1)Δt] [9]
which leads to the following simplification:
This expression indicates that the MFC at the time of the nth refocusing pulse can be estimated from the signal intensities of the preceding and succeeding echoes. As provided in equation [10], an echo train of N echoes can be used to obtain the MFC at N different times.
As an example of how to apply equation [10] to obtain an estimate for the MFC, the experiment described in Wismer may be considered. See Wismer et al., J Comput Assist Tomography 1988; I12:259-265, the entire disclosure of which is incorporated herein by reference.
1n[S1(ts)]≈1n[S1(0)]−2γ2ts2K(Δt). [11]
Thus, the signal intensity is preferably a linear function of ts2 on a semi-logarithmic plot. An estimate of the MFC at the time of the refocusing pulse can then be found from the slope of a line fitted to the data. In
Qualitatively similar behavior for the signal intensity as a function of ts has been reported for normal brain, bone marrow, lung tissue and microsphere phantoms. See the Hoppel, Rosenthal, Ganesan., and Stables Publications.
Referring to
One of the features of the present invention is the use of the RF pulse sequence. Referring to
According to one exemplary embodiment of the present invention, two pulse sequence programs implementing an ADSE may be used. A dual spin echo has the advantage over a single spin echo by simultaneously providing the information usable to estimate the MFC for two different time values. In one exemplary embodiment of the present invention, a conventional sequence with a single phase encoding step for each 90° radio frequency (RF) pulse may be used. For example, the sequence shown in
Using the exemplary techniques described in the present invention, the MFC may be estimated by applying equation [7] to the ADSE imaging data. For example, typical MFC values obtained in vivo for a human brain are using the methods and systems according to the present invention listed in Table 1. One prior in vivo human study, which allows for MFC estimates, gives data just for hemorrhagic brain lesions, which have a much higher MFC than healthy brain tissue. See Wismer Publication. This study, however, has been performed at a low field level of 0.6 T, and the MFC in normal brain was presumably too small to observe due to a small SNR. Since the MFC increases quadratically with the field strength, the MFC at 1.5 T is 6.25 times larger than at 0.6 T, its detection is easier.
Therefore, the exemplary embodiment of the system process and software arrangement according to the present invention is well-adapted to carry out the objects and attain the ends and advantages mentioned as well as those which are inherent therein. While the invention has been depicted, described, and is defined by reference to exemplary embodiments of the invention, such a reference does not imply a limitation on the invention, and no such limitation is to be inferred. The invention is capable of considerable modification, alteration, and equivalents in form and function, as will occur to those ordinarily skilled in the pertinent arts and having the benefit of this disclosure. The depicted and described embodiments of the invention are exemplary only, and are not exhaustive of the scope of the invention. Consequently, the invention is intended to be limited only by the spirit and scope of the appended claims, giving full cognizance to equivalence in all respects.
The present application claims priority from U.S. Patent Application No. 60/485,502, filed Jul. 8, 2003, the entire disclosure of which incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
Parent | 10564697 | Jan 2006 | US |
Child | 12710035 | US |