WO2005047916A1

WO2005047916A1 - Method and apparatus for the passive shimming of the fringe field of a superconductive magnet

Info

Publication number: WO2005047916A1
Application number: PCT/CA2004/001947
Authority: WO
Inventors: Juan M. Parra-Robles; Albert R. Cross; Giles E. Santyr
Original assignee: Parra-Robles Juan M; Cross Albert R; Santyr Giles E
Priority date: 2003-11-13
Filing date: 2004-11-12
Publication date: 2005-05-26

Abstract

Methods and devices for use in magnetic resonance imaging using a fringe field of the superconducting magnet. The fringe field of a superconducting magnet is shimmed or adjusted using passive shims to provide the homogeneity required for hyperpolarized gas magnetic resonance imaging. The magnetic field of the superconducting magnet is first mapped and measured. Given the desired homogeneity, shim calculations are performed to determine the configuration of the shim elements which would give the desired homogeneity. The optimization shim set is then positioned to result in a more homogeneous fringe field. The shim elements in the shim set may be steel bars as passive shim elements. Other active shim elements such a ferromagnetic materials acting as permanent magnets may be used in conjunction with the passive shim elements.

Description

METHOD AND APPARATUS FOR THE PASSIVE SHIMMING OF THE FRINGE FIELD OF A SUPERCONDUCTIVE MAGNET

Field of Invention

[0001] The present invention relates to magnetic resonance imaging and, more specifically, relates to methods and devices for use in reducing inhomogeneities in the fringe magnetic field of large and powerful magnets so that these fringe fields may be used for imaging purposes.

Background to the Invention

[0002] Shimming is the process of modifying the magnetic field produced by some field source in a manner that will produce a field with some desired characteristics. In magnetic resonance imaging or MRI, extremely uniform fields are required over a relatively large region. Typically only deviations as small as a few parts per million (ppm) can be tolerated for such fields. Even if designed for such homogeneity, magnet designs do not usually achieve such homogeneous fields due to construction inaccuracies, therefore additional corrections are required. These corrections are based on measurements of the field and the use of coil arrays of various geometries carrying certain currents (active shims) or locating ferromagnetic pieces (passive shims) of certain geometries and magnetic properties at different locations in the magnetic field. Usually, the field mapping is performed in such a way that the field distribution can be expressed in terms of spherical harmonic functions. Several coil sets are usually available for active shimming, each set producing a field correction approximating a relatively pure spherical harmonic. A similar effect is possible by using passive shims. [0003] Generally active shimming of the fringe field is impractical since extremely high currents are needed to correct the large field inhomogeneities present. These inhomogeneities can be in the form of strong approximately linear gradients. The power supplies for such an approach would be expensive and the current stability as well as the high power dissipation (i.e. heating) would be a significant concern.

[0004] Passive shimming of the fringe field is an attractive alternative, but must be done with care. The calculation of the shim set parameters is complicated by the strong variations of the field over the volume of the shim pieces. Bringing large pieces of ferromagnetic materials close to the magnet also poses safety concerns due to the strong magnetic forces exerted on the ferromagnetic pieces. It will also be undesirable to affect the homogeneity inside the magnet.

[0005] If the fringe fields of large magnets can be harnessed, it may be used for other types of imaging. Hyperpolarized Noble Gas (HNG) Magnetic Resonance Imaging (MRI) and Pre-polarized Magnetic Resonance Imaging (PMRI) provides exciting possibilities of using ultralow magnetic field strengths (<0.15 T) with reasonable sensitivity and increased contrast and immunity to susceptibility artifacts and other image degradation. The use of ultralow fields also reduces considerably the cost of the MR systems since the magnet is usually the most expensive component. Low field strengths also provide advantages for patient accessibility which is very attractive for many applications such as Interventional MRI. There is also increasing interest in performing Electron Paramagnetic resonance experiments at ultralow magnetic field to take advantage of the lower resonance frequencies.

[0006] The magnetic field outside of the superconductive magnets used in standard MRI machines provides an inexpensive and extremely stable magnetic field with different field intensities at different distances from the magnet. While the relatively low field strengths available in this field (fringe field) almost prohibits its use for most standard MRI applications due to the strong dependence of the sensitivity of the MR experiments on the field strength, sensitivity for HNG-MRI and PMRI is less dependent on the field strength and EPR takes advantage of the three orders higher magnetic moment of the electrons to make possible the use of relatively low field strengths.

[0007] A major obstacle for the use of the fringe field is the presence of extremely large field inhomogeneities mostly in the form of strong field gradients (typically 20-100 mT/m) the correction of which, using standard active shimming techniques, would require impractically high shim currents. Even in the case that such shim power supplies are available the temporal stability of these currents would be a major limitation.

[0008] There have been previous attempts at both passive and active shimming of magnetic fields. It should be noted, however, that the use of exclusively active shims is impractical since extremely high currents are needed to cancel the strong gradients present in the fringe field. If power supplies are available to provide such currents they would still be expensive, power consuming and, in most cases, the shim coils would need a cooling system. These power supplies could also be a source of field instabilities. [0009] The procedures of passive shimming within the magnet bore have been extensively reviewed in the literature. All procedures involve the use of steel pieces inside the magnet to correct the field inhomogeneities. The procedures involve positioning steel pieces in specific known positions inside the magnet or calculating the positions depending on what corrections are needed. Two relevant references are: - Romeo, F. and Hoult, D.I. "Magnet Profiling: Analysis and Correcting Coil Design" Magnetic Resonance in Medicine, 1, 44-65, 1984. - Hoult D.I. and Lee, D. "Shimming a superconducting nuclear- magnetic-resonance imaging magnet with steel", Review of Scientific Instruments, 56(1): 131-135, 1985.

[00010] A number of applications have been reported which make use of the field outside of the magnet bore for MR experiments. Diffusion experiments in the fringe field of superconductive magnets. These experiments make use of the strong gradients present in the fringe field of the superconductive magnet to measure very small diffusion coefficients. In this case no shimming is done since the large inhomogeneity in the form of a strong linear gradient is desired. These experiments are performed however in positions where the field is still very strong and very inhomogeneous (close to the magnet bore opening) and not suitable for imaging. Relevant publications in this area are: Kimmich, R. et al, "NMR Measurement of Small Self-Diffusion Coefficients in the Fringe Field of Superconducting Magnets", Journal of Magnetic Resonance, 91 , 136-140, 1991; and - Wu, D. and Johnson, OS. Jr. "Diffusion-Ordered 2D NMR in the Fringe Field of a Superconducting Magnet", Journal of Magnetic Resonance Series A, 116, 270-272, 1995. [00011] One interesting idea with respect to MRI in the fringe field of magnetic has been found. A method for MR imaging in the fringe field of a superconductive magnet has been patented (Cho, Z.H. and Wong Jr., E.K. U.S. Patent # 5,023,554, June 11 , 1991 ). This method proposed the use of specially designed pulse sequences to obtain magnetic resonance images in the highly inhomogeneous fringe field of a superconductive magnet. No attempt at shimming is done. This method, as well as the above referred diffusion measurements, have serious disadvantages which are discussed below. [00012] At the present, low field imaging use permanent and resistive magnets which are much more expensive than using the fringe field of an existing magnet. Also those approaches, suffer from temporal instabilities of the field. Permanent magnets are heavy and resistive magnets are power consuming and both need a cooling system.

[00013] It is therefore an object of the present invention to mitigate, if not overcome, the deficiencies of the prior art. A shimming or adjusting technique that adjusts the fringe field of a large magnet (such as a super conducting magnet) but does not require a large power supply nor heavy magnets would be advantageous.

Summary of the Invention

[00014] The present invention provides methods and devices for use in magnetic resonance imaging using a fringe field of the superconducting magnet. The fringe field of a superconducting magnet is shimmed or adjusted using passive shims to provide the homogeneity required for hyperpolarized gas magnetic resonance imaging. The magnetic field of the superconducting magnet is first mapped and measured. Given the desired homogeneity, shim calculations are performed to determine the configuration of the shim elements which would give the desired homogeneity. The optimization shim set is then positioned to result in a more homogeneous fringe field. The shim elements in the shim set may be steel bars as passive shim elements. Other active shim elements such a ferromagnetic materials acting as permanent magnets may be used in conjunction with the passive shim elements.

[00015] In a first aspect, the present invention provides an apparatus for use with external magnetic fields of a magnetic resonance imaging (MRI) device, the apparatus comprising a plurality of shimming elements arranged and positioned in said external magnetic fields to adjust characteristics of said external magnetic fields; a vessel containing material to be imaged using said external magnetic fields, said vessel being arranged and positioned in said external magnetic fields to use said characteristics of said external magnetic fields as adjusted by said shimming elements.

[00016] In second aspect, the present invention provides a method for adjusting a fringe magnetic field of a magnet, the method comprising: a) measuring said fringe magnetic field to determining a magnetic field gradient of said fringe magnetic field; b) determining an effect of at least one shimming element on said fringe magnetic field; c) determining if said at least one shimming element on said fringe magnetic field produces a desired fringe magnetic field; d) adjusting a number and/or position of at least one shimming element and repeating steps b)-c) until said desired fringe magnetic field is achieved. [00017] In a third aspect, the present invention provides a kit of parts for use with external magnetic fields of a main magnet of a magnetic medical diagnostic machine, the kit comprising: a plurality of shimming elements to be arranged and positioned in said external magnetic field to adjust characteristics of said external magnetic field; a vessel for containing material to be imaged using said external magnetic field after said characteristics of said external magnetic field have been adjusted by said shimming elements.

BRIEF DESCRIPTION OF THE DRAWINGS

[00018] A better understanding of the invention will be obtained by considering the detailed description below, with reference to the following drawings in which:

Figure 1 illustrates a coordinate system that shows the various angles and points used in the calculations for the invention;

Figure 2 illustrates one embodiment of the invention; Figure 3 illustrates an end view of the embodiment of the invention illustrated in

Figure 2;

Figure 4 illustrates a coordinate system that shows the relevant angles and points used in the calculations involving a passive shim element;

Figure 5 illustrates the placement of the passive shim elements for one implementation of the invention;

Figure 6 illustrates the placement of the passive shim elements for another implementation of the invention;

Figure 7 is a graph illustrating the inhomogeneities of a fringe magnetic field both before and after implementing one aspect of the invention;

Figure 8 is a graph illustrating the inhomogeneities of a fringe magnetic field for another implementation of the invention both before and after the fringe field has been corrected;

Figure 9 illustrates the hyperpolarized¹²⁹ Xe (HXe) spectrum after passive shimming for one implementation of the invention;

Figure 10 illustrates the HXe spectrum for the implementation referred to in

Figure 9 after both active and passive shimming of the magnetic field;

Figure 11 illustrates the HXe spectrum after passive shimming for another implementation of the invention; and

Figure 12 illustrates the HXe spectrum for the implementation referred to in

Figure 11 after both active and passive shimming.

Detailed Description

[00019] To map and calculate the fringe magnetic field of a large magnet and to find the number of shim elements required the following method may be used with reference to Figure 1. It should be noted that the axes provided on the figures are for references purposes and are identified in a conventional manner. [00020] The scalar potential at point P(r) produced by an infinitesimal volume dV of shim element, of susceptibility , located at position Q(f') is given by,

where k is the unit vector in the z direction, r and Ψ are position vectors for points P and Q respectively and V_ρ is the gradient with respect to Q. To obtain the total potential Φ , in the presence of a shim element, this equation must be integrated over the volume of the element V_elem giving:

[00021] Taking into account that H = -V_p Φ and changing the order of the derivative and integral operations, the magnetic field in the z direction is given by:

Expanding . in spherical harmonics as: |F - r'\ ^∞s[^m(Φ - Ψ⁾] (4)

where P_n,_m are the associate Legendre functions of order n and degree m; ε_m =l for m = 0 and ε_m = 2 for m ≠ 0. Using the properties of spherical harmonics to find the derivatives in Eq. (3) and taking the sums out of the integral, the following expression can be obtained: 1 °° " r I H_z(r,θ,φ) = - -∑∑A_1,mrⁿP_n, ∞sθ) [∞sm H_{n m} +smmφ K_{n m}\ (₅) ^π „=o m=0 where: (n - m + 2)\ A.. = -ε„ (n + m)l

^Hn,_m = X (∞* <x)∞* >n ψ ^dV (6.1 )

^κ _n,m = X dV (6.2)

[00022] If more than one shim element is used, the principle of superposition applies and for the general case where the shim elements have different magnetic susceptibilities, the expression for the field coefficients H_{n m} and K_{n m} can be written as:

^Hn,m ⁼ Σ ^Hn,m (⁷-1 ) k

where H*_m and

are the coefficients corresponding to the k^th shim element of susceptibility χ_k , calculated using Eqs. (6.1) and (6.2). [00023] For shimming within the bore of the magnet, where the field is relatively constant over the volume of the shim elements, H_z(r') is taken out of the integrals in Eqs. (6.1) and (6.2). However, for shimming in the fringe field, these integrals must be numerically computed from the values of the magnetic field H_Z(F') over the volume of the shim element. Once the values of the coefficients H„ _m and K_{n m} have been found, the field strength at every point in the volume of interest can be computed.

[00024] Referring to Figures 2 and 3, an embodiment of one aspect of the invention is illustrated. A superconducting magnet 10 has a magnetic field 20. A device 30 is in the fringes of the magnetic field 20 and contains a sample for imaging. The sample (with an RF coil coiled around it to and referred generally as 40) is RF shielded using shielding 50. The fringe magnetic field is shimmed or adjusted using passive shim elements 60 contained inside the fringe magnetic field. For hyperpolarized gas applications, the polarization and flow system for the gas is in box 80 and the flow of gas (possibly ΗXe) is toward the device 30 by was of conduit 90.

[00025] In one implementation, the shimming procedure was used to passively shim the fringe field at the 8.5 mT and 17 mT positions. The first position was located 120 cm from the centre of the superconducting magnet and was selected such that after shimming a field strength of 8.5 mT would be available over a 2 cm diameter spherical volume (DSV). The second position was selected closer to the centre of the superconductive magnet (90 cm) in order to obtain a field strength of 17 mT (after shimming) over a 6 cm DSV. The selection of those positions was based on an initial field mapping of the fringe field (as discussed below). The second position was expected to yield higher SNR and be more useful for in vivo (i.e. animal) imaging purposes.

[00026] In an implementation that involved an HXe polarization system for imaging, the ultra-low field MR imaging system used the fringe field of a 30 cm bore superconductive magnet (1.89 T, Magnex, Exon, England) which permitted field strengths up to 20 mT at the surface of the magnet's dewar. Imaging was accomplished using a 26 cm diameter gradient and shim set (Bruker B-GS 30/C- 19, Ettlingen, Germany) powered by the gradient and shim power supplies of the 1.89 T system (Techron 7700 and Resonance Research MXA-18/4V 0, respectively) and controlled by an MRRS (Surrey, UK) MR5000 console. The configuration for this system is similar to that in Figures 2 and 3. [00027] The electronics were based on a polarimeter design proposed by

Saam (B.T. Saam, M.S. Conradi, "Low frequency NMR polarimeter for hyperpolarized gases", J. Magn. Resosn. 134, 67-71 (1998). Modifications to this design included quadrature phase detection and interfacing to the MR5000 console. For the experiments at 8.5 mT, a solenoidal RF coil with a bore diameter of 10 mm, length of 12 mm (560 turns of 34 AWG coated copper wire) and tuned to 100 kHz was used as a field mapping probe and for homogeneity measurements. At 17 mT, a solenoidal RF coil of similar dimensions but with only 250 turns and tuned to 200 kHz was used for field mapping. A larger RF coil was used to obtain signals for the final active shimming and homogeneity measurements over a larger volume at 17 mT. This coil was a split solenoid of diameter 4.5 cm, length 6 cm and 1 cm separation between the two winding sets (each 25 mm wide and made of 50 turns of 22 AWG coated copper wire). [00028] Hyperpolarized natural abundance xenon gas (26.4% ¹²⁹Xe) was produced continuously from a gas mixture (1% Xenon, 10% Nitrogen and 89% Helium) using a flow polarization system described previously in I.L. Moudrakovski, S. Lang, CI. Ratcliffe, B. Simard, G. Santyr, and J. Ripmeester, Chemical Shift Imaging with Continuously Flowing Hyperpolarized Xenon for the Characterization of Materials, J. Magn. Reson. 144, 372-377 (2000) The system used circularly polarized light from a 60W diode array laser (λ = 794.8 nm , Coherent, Santa Clara, USA) and produced xenon polarizations of up to 20% at a gas flow rate of 3.5 cc/min (at room temperature and 1 atm. pressure). [00029] To adjust the fringe magnetic field, the gradient/shim coil set was positioned with its axis (z') parallel to the magnet's axis (i.e., aligned with the magnetic field direction) and slightly lower along the y axis (see Fig. 3) due to space restrictions in the magnet room. These restrictions were imposed by the presence of the xenon polarization system and walls. Also due to these restrictions, the two positions were located on different sides of the magnet (right for 8.5 mT and left for 17 mT).

[00030] The gradient/shim set was rotated about this axis by an angle such that the x' direction of the gradient/shim set pointed towards the magnet centre as shown in Fig. 3. This choice of geometry simplified the calculations by making the field variations symmetric about the x' axis. Acrylic end pieces located between the two layers of the RF shield were used to hold the shimming elements (see Fig. 3).

[00031] Initial field mapping was performed using a Hall-effect gaussmeter

(FW Bell 640, Orlando, FL, USA) and a simple probe positioning system consisting of a circular piece of acrylic, containing an array of precision holes in the x'y' plane to accommodate the probe, that could be moved along the axis of the gradient/shim set to different z' positions. Once the field 20 was passively shimmed (as described below) to a level that permitted HXe signals to be observed directly, field mapping was performed more rigorously using continuously flowing HXe gas through the RF coil 40 which was moved to different positions within the gradient/shim set and enabled precision field mapping in a relatively short time (1-2 min per measurement point). [00032] Commonly available construction steel rods of diameters ranging between 1 and 3 cm were used as shim elements. As a first step, the magnetic susceptibility χ of the steel was experimentally measured by mapping of the effect of one pair of steel rods on the fringe field and comparing the generated spatial field harmonics to those predicted for equivalent steel rods of similar dimensions and susceptibility equal to one. The magnetic susceptibility of the steel was estimated to be 10.9 ± 2.1 , which is similar to reported susceptibility values for cold rolled steel.

[00033] The initial estimate for the shim set parameters was obtained by assuming the field strength to be constant over the rod and using the standard passive shimming calculation technique. The number of steel rods needed to produce the desired correction of the field gradient depended on the lowest order harmonic introduced that would be tolerated for the desired target field homogeneity. This order is usually determined by the size of the desired volume of homogeneity. However, for this implementation, this condition was relaxed to allow harmonics to be introduced up to the point where they were correctable by the corresponding active shim in order to use as little passive shimming material as possible.

[00034] The optimum arrangement of shim elements to produce the desired correction was obtained by iteratively adjusting the following parameters: radius R , distance from the centre r (of the magnet) and length L (L = 2r /tanα ), from the initial estimate and numerically calculating H_z(r,θ, f using Eqs. (5) and (6), and the measured field maps. For simplicity, the angular position Ψ was not changed from the initial estimates, but could be varied as well. Figure 4 is provided as a reference and it illustrates a passive shim element 60. [00035] This calculation procedure was implemented using the optimization toolbox (simple search minimization algorithm) in Matlab 6.5 (The Mathworks, Natick, MA, USA) and the field coefficients to be minimized were expressed in the form:

[00036] After correcting most of the field inhomogeneities using the steel rods, the field homogeneity was improved further using the active shims, which were systematically adjusted to provide the narrowest line-shape. The linewidth was estimated by fitting the spectra to a Lorentzian function. Spectra were obtained by releasing the HXe gas (in continuous flow mode) into imaging phantoms: a 10 mm diameter open glass cell and a 3.7 cm diameter hollow plastic ball (which approximates the rat lung volume).

[00037] In the above implementation, there were two sets of results - one for the 8.5 mT position and one for the 17 mT position.

[00038] For the 8.5 mT position, the shimming procedure resulted in a set of two 1.2 cm diameter steel rods with dimensions given in the table below (one pair of passive shimming rods) and Figure 5. The x'-axis was rotated 54° clockwise about the z direction from the x-axis to make field variations symmetric about the x'-axis. At the 17 mT position, the optimum shim set consisted of two pairs of steel rods (see table below (two pairs of passive shimming rods) and Fig. 6, with the x-axis rotated 16° counterclockwise.

[00039] For clarity, it should be noted that the entries for the initial estimates are for results obtained using known procedures for shimming the field inside the bore of the magnet. The optimized design entries are the results obtained when one aspect of the present invention provided better results in terms of homogeneity than the known procedures.

[00040] As can be seen in Figure 5, the shim elements 60A, 60B are each at 22.5° from the positive x'axis with r₀ being the distance from the center of the magnet to the center of the shim elements 60A, 60B.

[00041] For Figure 6, there are two sets of passive shim elements, 60C and

60D. Shim elements 60C are at 15° to the positive x'axis and are ri distance from the center of the superconductive magnet. Shim elements 60D are at 75° to the positive X' axis and are at r₂ distance from the center of the superconducting magnet. As can be seen from the table below, the three sets of shim element pairs each have varying lengths and radii.

[00042] Figure 7 shows the results of the field mapping of the fringe field along the radial direction ( ^ at the 8.5 mT position. A strong approximately linear gradient (30 mT/m) was found to be the main contributor to the field inhomogeneity.

[00043] The results of the field mapping at the 17 mT position are shown in

Figure 8. At this position, a strong field gradient (92 mT/m) was found to be the main contributor to the field inhomogeneity. Figures 7 and 8 also show the field maps for both positions after field correction using the passive shims. [00044] At 8.5 mT, the strong field gradient (30 mT/m) was effectively removed using the passive shims improving the field homogeneity to 0.1% over the 2 cm DSV from the initial 7.2%The shim set design that produced the correction at 8.5 mT is significantly different from the initial estimate (see table above). The rods are 1.3 cm shorter (20.7 cm). This difference is due to the relatively strong variation of the field strength over the length of the rod, mainly in the form of a second-order zonal harmonic H₂,o- As a result, the initial estimate does not minimize the -/₂,o harmonic and would have resulted in the introduction of a relatively large inhomogeneity (6.35 x 10^"4 mT/cm²). The optimized shim set produced a H₂,o harmonic two orders of magnitude smaller (6.70 x 10^"6 mT/cm²) while producing a Hι_tι harmonic (the desired field correction) larger (-31 mT/m) than the one produced by the initial estimate design (-27 mT/m).At 17 mT, the very strong field gradient (92 mT/m) was canceled (Fig. 8) by the passive shims thereby increasing the field homogeneity to 0.2 % over the 6 cm DSV from the initial 25%. In this case (two pairs of shimming rods), due to the lack of cylindrical symmetry with respect to the centre of the volume of interest, the field strength at the positions of the two pairs of rods is different. The initial shim estimate introduces second (H₂,₂=1.12 x 10^"2 mT/cm²) and third (H₃,₃=1.12 x 10^"2 mT/cm³) order harmonics while the optimized design, by using pairs of rods with different diameters and located at different distances from the center of the volume of interest, compensates for the field asymmetry and produces much smaller second (H₂,₂=2.02 x 10^"5 mT/cm²) and third (H₃,3=9.22 x 10^"4 mT/cm³) order harmonics. Similarly in the case of one pair of rods, the present invention uses shorter rods which minimizes l+2,o and maximizes the desired Hι,ι correction [00047] Figure 9 shows the HXe spectrum obtained at 8.5 mT after passive shimming (linewidth approximately 100 Hz). Figure 10 shows the spectrum after both passive and active shimming (linewidth: -20 Hz). Figure 11 shows the HXe spectrum (linewidth: -200 Hz) obtained at 17 mT after passive shimming. Figure 12 shows the spectrum after both passive and active shimming (linewidth: -30 Hz). The linewidths obtained are better than the target homogeneity of 40 Hz, which was desired for the imaging experiments. It should be noted that no signals were observable in either case without the passive shims. [00048] Based on the above, the present invention provides a method to shim the fringe field of a superconducting magnet in order to produce a homogenous region for hyperpolarized gas imaging. [00049] The practical implementation of the shimming procedure at two positions in the fringe field of the superconducting magnet produced the desired volumes of homogeneity at 8.5 mT (2 cm DSV) and 17 mT (6 cm DSV), sufficient to allow HXe signals to be obtained (Figs. 9 and 11 ), thus permitting improved homogeneity by active shimming using these signals. The final field homogeneity (approximately 0.017%) allowed reasonable HXe spectra to be obtained (Figs. 10 and 12) with a linewidth of 19±3 Hz at 8.5 mT and 33±6 Hz at 17 mT. Further homogeneity improvements were not possible due to the lack of fourth order zonal harmonic (H₄₀ ) correction in the active shim system used.

[00050] It should be noted that though the final homogeneity (0.02% or 200 ppm) seems large compared to the typical field homogeneity in clinical systems (<10 ppm), this is a relative quantity, which depends on the static magnetic field strength. For imaging purposes the homogeneity requirement is that the linewidth produced by the field inhomogeneities must be smaller than the desired pixel size (in frequency units). A homogeneity of 200 ppm at 17 mT (0.0034 mT or 40 Ηz for ¹²⁹Xe) is comparable to a homogeneity of 1.5 ppm at 1.89 T (0.0030 mT or 35 Ηz for ¹²⁹Xe), which is quite acceptable.

[00051] Though the differences between the optimized shim sets and the initial estimates can be as small as a few millimeters, they strongly affect the field correction produced by the shim sets, resulting in differences of several orders of magnitude. Consequently, the shim sets have to be accurately constructed. It is preferable that the shim holder allows for fine positioning (accurate to within 0.1- 0.5 mm) of the shimming elements. As well, this is just one approach to optimizing the shim set design and others are possible including varying χ and/or Ψ.

[00052] The presence of the steel rods in the fringe field did not greatly affect the homogeneity inside the superconducting magnet. The largest estimated field inhomogeneity introduced was about 5 ppm, which is easily correctable by the active shims of the high field system. The forces exerted on the steel rods by the magnetic field were small enough that no anchoring for the system was needed. At higher fields or when using larger amounts of steel, this factor may require careful consideration.

[00053] The shimming or adjusting process can be summarized as a three step procedure. The steps are:

[00054] Field Mapping: In this step the magnetic field in the region of interest is measured using a proper measuring device. The field is initially measured at a number of positions in a volume of interest at certain distances from the centre of the superconductive magnet. If a certain field intensity is desired then an initial field mapping is necessary to obtain the optimum location of the system before the final field mapping for shim calculation. In this initial mapping, the field intensity is measured over a relatively large range of positions in a direction radial from the center of the magnet. This way a first estimate of the magnetic field gradient is obtained. Based on this estimate and the desired volume of uniformity and the magnetic properties of the ferromagnetic material to be used, the position of the center of the region of interest is calculated. If the magnetic susceptibility of the ferromagnetic materials to be used is unknown it can be estimated from an additional experiment at this stage by locating some pieces of magnetic material in certain positions and re-measuring the field. The changes found will allow the estimation of the susceptibility. This initial mapping can be extended to a full 3D mapping if desired, to obtain further information for the proper positioning of the system. At the selected position, the field is measured using a positioning system that will allow a very accurate (error < 1 mm) positioning of the probe of the measuring device. The positions at which the field has to be measured must be sufficient to allow the expansion in spherical harmonics up to the degree necessary to produce the desired field homogeneity. Expansion using other orthogonal base functions is also possible. Unlike standard shimming techniques, the range of the measurements has to be much larger than the volume of interest, to provide accurate information about the distribution of field intensities over the positions where the shim pieces are expected to be. These positions can be estimated approximately from the initial mapping data.

[00055] Shim Calculations: Based on the size of the volume of interest

(VOI), the desired homogeneity and the field maps obtained in the previous step, the optimum shim design is calculated by numerically integrating the effect of the ferromagnetic material on the fringe field for different positions and geometries of the ferromagnetic materials. Usually rods of ferromagnetic materials are used in pairs as shimming elements due to symmetry considerations (see Appendix 1), but different geometries can also be used. The number of shim elements is determined by the degree of the highest harmonic that is to be avoided of been introduced by the shim elements. The positions and characteristics of the passive shims are then numerically optimized to produce the desired correction with the minimum amount of ferromagnetic material. If very high precision is needed the designed shim set can be placed in position and new field maps obtained to further optimize the design in an iterative process. This calculation-optimization procedure differs from the standard shimming technique by effectively accounting for the significant variation of the field strength over the volume of the shimming elements.

[00056] Final Shimming: The optimized shim set is located in its position and a more accurate field map is obtained preferably using a MR probe. This map will allow the verification of the effectiveness of the passive shimming and provide initial data with which to perform a finer shimming using active shims. ( [00057] It should be noted that the above described methods and devices can be used for other purposes and that other materials may also be used. The device described above, with some modifications, can be used for hyperpolarized gas imaging of human body parts such as lungs. It may also be used for purposes such as blood perfusion work, EPR, pre-polarized magnetic resonance

(MR) imaging, and conventional proton MR imaging. Other media may also be used other than hyperpolarized gas. Other hyperpolarized contrast media, such as noble gases or carbon 13 may also be used. [00058] Regarding the materials used, the shimming elements may be passive or active. Any material which may affect a magnetic field, such as ferromagnetic materials, may be used to construct the shimming elements. The fringe magnetic fields of the superconducting magnet may be adjusted using passive shimming elements as described above or by using smaller active shimming elements that do not necessitate the use of high current sources. These external magnetic fields of the main magnet may be adjusted by using a combination of active and passive shimming elements using the technique outlined above. While the above description relates optimizing the number, placement, and positioning of the shimming elements, optimization is not required to achieve useful results. Useful images may also be obtained using non- optimized placement and positioning of shimming elements. [00059] While the above embodiment is described and illustrated as being of a cylindrical design, other designs for the shim holder and for the positioning of the shim elements can be used. The shim elements may be included in the construction of other components of the MR system such as the shim/gradient tube, RF coils, or may be placed and positioned in an independent holder. The shim elements may be, as illustrated in the figures, positioned as part of the vessel which contains the material to be imaged.

[00060] The above-described apparatus may be embodied in a kit of parts to be used to construct an accessory system to a medical diagnostic machine such as an MRI machine. The adjustable number and positioning of shimming elements and of the placement of the homogenous volume produced by this method will allow the movement of the apparatus to different positions in the fringe field to obtain different field strengths for MR imaging of objects. In one contemplated embodiment, a human subject could be imaged while in an upright position. This will allow useful images of the subject's lungs to be obtained. [00061] It should further be noted that while the method described above is used to adjust the homogeneity of the fringe magnetic field, other characteristics of this external magnetic field may also be adjusted using the same technique. The method described above may be used to adjust the external or fringe magnetic field of any main magnet. The fringe field could also be modified prior to the use of the above invention to extend the capabilities of the system. This way, increased field strengths and improved symmetry of the fringe field may be made available. It has been found that the above technique is most suited for medical imaging uses.

[00062] The above described implementation may be extended by taking into account the small components of the field that are not parallel to the Z direction. This can be done by accounting for these components in the calculations. When included in the calculations, specially shaped shims should result that compensate for these components.

[00063] A person understanding this invention may now conceive of alternative structures and embodiments or variations of the above all of which are intended to fall within the scope of the invention as defined in the claims that follow.

Claims

We Claim:

1. Apparatus for use with external magnetic fields of a magnetic resonance imaging (MRI) device, the apparatus comprising: - a plurality of shimming elements arranged and positioned in said external magnetic fields to adjust characteristics of said external magnetic fields; - a vessel containing material to be imaged using said external magnetic fields, said vessel being arranged and positioned in said external magnetic fields to use said characteristics of said external magnetic fields as adjusted by said shimming elements.

2. Apparatus according to claim 1 wherein said plurality of shimming elements are arranged and positioned using a method comprising: a) measuring said external magnetic fields to determine magnetic field inhomogeneities of said external magnetic fields; b) determining an effect of said shimming elements on said external magnetic fields; c) determining if said shimming elements on said external magnetic fields produces a desired external magnetic field; and d) adjusting a number and/or position of said shimming elements and repeating steps b)-c) until said desired external magnetic field is achieved.

3. Apparatus according to claim 1 wherein said at least one of said shimming elements is a passive shimming element.

4. Apparatus according to claim 1 wherein at least one of said shimming elements is an active shimming element.

5. Apparatus according to claim 1 wherein said plurality of shimming elements is constructed of ferromagnetic materials.

6. Apparatus according to claim 1 wherein one of said characteristics is homogeneity.

7. Apparatus according to claim 1 wherein said external magnetic fields are used for imaging using a hyperpolarized contrast medium.

8. A method for adjusting a fringe magnetic field of a magnet, the method comprising: a) measuring said fringe magnetic field to determining a magnetic field gradient of said fringe magnetic field; b) determining an effect of at least one shimming element on said fringe magnetic field; c) determining if said at least one shimming element on said fringe magnetic field produces a desired fringe magnetic field; and d) adjusting a number and/or position of at least one shimming element and repeating setps b)-c) until said desired fringe magnetic field is achieved.

9. A method according to claim 8 wherein step d) is repeated until said fringe magnetic field has a desired homogeneity

10. A method according to claim 8 wherein at least one of said at least one shimming element is passive

11. A method according to claim 8 wherein at least one of said at least one shimming element is active.

12. A method according to claim 8 wherein steps b) and c) include determining magnetic field strength over a volume of each one of said at least one shimming elements.

13. A method according to claim 8 wherein said desired fringe magnetic field is used for imaging using a hyperpolarized contrast medium.

14. A method according to claim 8 wherein said magnet is a superconducting magnet.

15. A kit of parts for use with external magnetic fields of a main magnet o a magnetic medical diagnostic machine, the kit comprising: - a plurality of shimming elements to be arranged and positioned in said external magnetic field to adjust characteristics of said external magnetic field; and - a vessel for containing material to be imaged using said external magnetic field after said characteristics of said external magnetic field have been adjusted by said shimming elements.

16. A kit according to claim 15 wherein said external magnetic fields are fringe magnetic fields of said main magnet.

17. A kit according to claim 15 wherein said material to be imaged comprises a body part of a human being.

18. A kit according to claim 15 wherein a placement and arrangement of said plurality of shimming elements is adjustable.

19. A kit according to claim 15 wherein said main magnet is a superconducting magnet.

20. A kit according to claim 15 wherein said at least one of said shimming elements is a passive shimming element.

21. A kit according to claim 15 wherein at least one of said shimming elements is an active shimming element.

22. Apparatus according to claim 1 wherein at least one of said shimming elements is a previously magnetized material.

23. A method according to claim 14 wherein said magnet is a superconducting magnet of a clinical magnetic resonance system.

24. A kit according to claim 19 wherein said main magnet is a superconducting magnet of a clinical magnetic resonance system.

25. A kit according to claim 15 wherein said apparatus is movable to different positions in said external magnetic fields to obtain different field strengths for magnetic resonance imaging.

26. A kit according to claim 25 wherein said different field strengths are obtained in a homogenous volume obtained using the method of claim 8.

27. A method according to claim 8 wherein said desired fringe magnetic field is used for a magnetic resonance technique which requires a region of homogeneous magnetic field.

28. A method according to claim 27 wherein said magnetic resonance technique is EPR.

29. A method according to claim 27 wherein said magnetic resonance technique is nuclear magnetic resonance.

30. A method according to claim 27 wherein said magnetic resonance technique is pre-polarized magnetic resonance.