WO1992004640A1

WO1992004640A1 - Method of obtaining images representing the distribution of paramagnetic material in solution

Info

Publication number: WO1992004640A1
Application number: PCT/GB1990/001379
Authority: WO
Inventors: David John Lurie; James Macdonald Strachan Hutchison; John Rowland Mallard
Original assignee: British Technology Group Ltd
Priority date: 1990-09-06
Filing date: 1990-09-06
Publication date: 1992-03-19

Abstract

In the known method of obtaining image information representing the distribution of paramagnetic material in solution, which comprises the steps of applying radio-frequency radiation to excite EPR resonance in the solution while obtaining an NMR image signal of selected nuclei of the solvent, thereby to enhance the signal obtained from those selected nuclei which interact with electrons excited by the radio-frequency radiation, the invention provides the improvement of using, as the method of obtaining the NMR image signal, the magnetic field cycling NMR method (in which each cycle includes a polarisation period, an evolution period and a detection period) and performing the EPR resonance excitation during the evolution period (in which the NMR magnetic field is of reduced strength). This enables a low EPR resonance excitation power to be used without requiring a reduction in the signal-to-noise ratio of the NMR image signal obtained.

Description

METHOD OF OBTAINING IMAGES REPRESENTING THE

DISTRIBUTION OF PARAMAGNETIC MATERIAL IN SOLUTION

This Invention relates to a method of obtaining Images representing the distribution of paramagnetic material In solution, including free radicals. The Invention has application inter alia in the determination of the distribution of free radicals, which may have been previously injected, in living tissue.

Electron paramagnetic resonance (EPR) imaging using methods analogous to those employed in nuclear magnetic resonance (NMR) imaging but at much higher nutation frequencies is known. However, in view of the strong magnetic field gradients necessary to achieve good spatial resolution, EPR imaging has so far been restricted to small samples.

It 1s also known that if nuclei of a solvent in which paramagnetic material is dissolved are excited to nuclear magnetic resonance and the NMR resonance is observed, a dramatic enhancement of the observed NMR signal may be obtained if the paramagnetic material is simultaneously excited to EPR resonance. This phenomenon may be utilised to obtain image information regarding the spatial distribution of paramagnetic material 1n solution, and our US Patent No. 4891593 (the contents of which are incorporated herein by this reference to it) discloses a method of obtaining image information representing the distribution of paramagnetic material in solution which comprises the steps of applying radio-frequency radiation to excite EPR resonance in the solute and obtaining an NMR image signal of selected nuclei (preferably protons) of the solvent, the signal from those selected nuclei which interact with electrons excited by the rf radiation being enhanced. As will be understood, the solution to which this method is applied may be inhomogenously distributed 1n a sample, of living tissue, for example, and the enhanced parts of the obtained NMR signal then represent the spatial distribution of the paramagnetic material within the tissue sample. The NMR signal enhancement, which occurs in regions of the sample where paramagnetic material is present and influences the NMR proton relaxation rate, and which appears in the final image as a locally increased Intensity of the image, may be defined as E = A_z/A₀, where A_z and A₀ are the NMR signals with and without EPR irradiation. It is known that E depends on the concentration of the paramagnetic solute and on the square of the EPR irradiation radlofrequency magnetic field which 1s in turn proportional to the power of the EPR irradiation. If the conductivity of the sample is assumed to be constant, the power of the EPR irradiation required to produce a given value of E can be shown to be proportional to the square of the EPR irradiation frequency.

Since a proportion of the applied radlofrequency power is always absorbed by the sample, it is desirable to minimise the applied power while maintaining a detectable enhancement. Thus 1t 1s desirable to use as low a value of the polarising magnetic field B₀ as possible, since the EPR frequency Is proportional to the B₀ NMR magnetic field. Considering the signal-to-noise ratio (SNR) of the NMR image, however, it is known that the SNR decreases rapidly with decreasing B₀. Thus the value of B_Q should be maximised to optimise the SNR, which will in turn improve the sensitivity of detection of the paramagnetic species. Thus there are, apparently, conflicting requirements applying to selection of the value of B₀.

It is an object of the present invention to provide a method of obtaining image information representing the distribution of paramagnetic material in solution, of the kind outlined above, in which these apparently conflicting requirements as to the magnitude of the NMR magnetic field B_Q can both be satisfied and the apparent conflict is resolved. This object is achieved, according to the invention, by obtaining the NMR signal by use of the NMR method known as magnetic field cycling NMR.

The magnetic field cycling NMR method, which is known as a method of studying NMR relaxation and other phenomena at extremely low field strengths (F. Noack, Prog. NMR Spectrosc. JJL. 171(1986)), has three distinct periods during each of which B₀ has a different value: polarisation at B_QP (high field), evolution at B₀ ^e (low field) and detection at B₀ ^d (intermediate or high field).

According to the present invention, therefore, there is provided a method of obtaining Image information representing the distribution of paramagnetic material in solution in a solvent, comprising the steps of obtaining an NMR image signal of selected nuclei of the solvent by the magnetic field cycling NMR method in which each cycle includes a polarisation period followed by an evolution period followed 1n turn by a detection period, and only during the evolution period of each cycle, applying radio-frequency radiation at a frequency appropriate to the magnetic conditions then appertaining so as to excite EPR resonance in the paramagnetic material, thereby to enhance the NMR signal obtained from those selected nuclei which Interact with electrons of the paramagnetic material excited by the radio-frequency radiation.

The method according to the invention will be more fully explained in the following description with reference to the accompanying drawings, In which:- Figure 1 is a representation of the magnetising coils and an NMR transmit/receive coll of a whole-body NMR scanning apparatus of known kind, with the addition of an EPR resonator:

Figure 2 shows, on a larger scale, the NMR transmit/receive coil and EPR resonator shown in Figure 1, together with a container for a sample which is to be examined; and Figure 3 is a chart showing the relative timings of events during one cycle of the field cycling NMR method applied by means of the apparatus shown in Figures 1 and 2 while carrying out the method according to the invention. As shown in Figure 1, whole-body NMR scanning apparatus of known kind comprises four coaxially arranged colls for generating a steady B₀ vertical magnetic field, namely a large main pair of colls 11 and a smaller pair of outer coils 12, together with gradient coils (not shown in detail) which are located in known manner on a gradient coil tube 13 of which the axis 1s perpendicular to that of the coils 11 and 12. In order to adapt this known arrangement for performing field-cycling NMR there are provided a further pair of smaller coils 14 coaxial with the coils 11 and 12 and located symmetrically between the colls 11. At the centre of the apparatus, i.e. where the axis of the gradient coll tube 13 intersects that of the coils 11 and 12, a transmit/receive NMR signal coll 15 is located coaxial with the gradient coil tube 13, and, as shown in Figure 2, a tube 16 for containing a sample which is to be examined is disposed coaxially within the coil 15. Surrounding the tube 16 is an EPR resonator 17 consisting of twenty loops connected in parallel to energising leads 18 connected to an EPR excitation signal generator (not shown) comprising a synthesised microwave frequency generator driving a broadband amplifier to whose output the leads 18 are connected. Suitable trimming capacitors (not shown) are included for tuning the resonator 17 to the appropriate EPR frequency and matching it to the leads 18. The illustrated resonator 17 could, if desired, be replaced by a so-called birdcage resonator, of known kind. In a particular embodiment of apparatus as above described which has been used for carrying out the method according to the invention, the coils 11 and 12, which are energised so that their fields are additive, were arranged to provide a steady B₀ magnetic field of 0.01 T at the centre of the apparatus, giving a proton NMR frequency of 425kHz. The coil 15, with a diameter of 85mm, was used for transmission and reception at that frequency, being connected to a suitable RF transmitter and receiver via a passive transmit/receive switch. The coils 14, which are connected electrically in series and wound so that their magnetic fields are additive (but opposed to that generated by the coils 11 and 12, as explained below), were of air-cored water-cooled type with an internal diameter of 220 mm and an inductance of 24 mri. Each of the coils 14 had 188 turns of 2.5 mm diameter copper wire and, energised with a current of 3.67A, they produced a magnetic field of 0.005 T at the centre of the apparatus. They were energised at a selectable constant current by a constant-current power supply, and the current was switched on and off, as explained below, using M0SFET transistors under the control of the pulse programmer of the NMR apparatus. The switching time was less than 10 ms.

In carrying out the method of the invention with the above described apparatus, the NMR apparatus is operated in the field cycling mode in which each cycle comprises a polarisation period followed by an evolution period followed in turn by a detection period. With the tube 16 containing a sample which is to be examined and which includes a possibly inhomogenously distributed solution of paramagnetic material in a solvent which contains hydrogen nuclei (protons), the nuclear magnetisation of the protons is allowed to build up during the polarisation period under the influence of the B₀ magnetic field of 0.01 T produced by the colls 11 and 12 alone, i.e. with the coils 14 unenergised. Then the coils 14 are energised to generate a filed opposed to that generated by the coils 11 and 12 so that (as shown by the top line of Figure 3) the net magnitude of the B₀ field quickly falls from 0.01 T to (in a particular case) only 0.0051 T, at which level it then remains during the evolution period of the cycle.

During this evolution period, as shown 1n the second line of Figure 3, an EPR excitation signal is applied to the EPR resonator 17. In a particular case, where the sample being examined was a phantom containing a 2 mM aqueous solution of the nltroxide free radical TEMP0L (4-hydroxy-2,2,6,6-tetramethyl- piperidlne-i-oxyl) at room temperature, the EPR irradiation frequency was fixed at 160 MHz at a power level of 1 watt, and the examination was carried out by irradiating one of the characteristic EPR lines of the nitroxide triplet which were observed at B₀ values of 0.0037 T, 0.0051 T and 0.0072 T, these values of B_Q ^e being obtained by suitable selection of the constant current applied to the colls 14. The intermediate resonance was used for most experiments.

The evolution period is then terminated by switching off both the EPR resonator and the colls 14, so that the B₀ field quickly resumes its value of 0.01 T for the ensuing detection period of the cycle. Alternatively, the current supplied to the coils could be reduced, rather than switched off completely, so as to provide during the detection period a field B₀ ^d greater than the evolution period field B₀ ^e but still less than the field B_QP provided during the initial polarisation period. During the detection period, in normal NMR manner, a radio frequency NMR signal is applied to the coil 15 (as shown by the third line in Figure 3) and gradient field signals are applied to the gradient coils (as indicated by the next three lines of Figure 3). This results in an output NMR image signal (represented in the seventh line of Figure 3) which is detected by the coil 15. As already mentioned, parts of this signal which are due to protons which have interacted with adjacent excited paramagnetic material are enhanced, so that the corresponding parts of the final image obtained are also of enhanced intensity and the final image then indicates the distribution of excited paramagnetic material in the sample being examined. It will be understood that, by carrying out the EPR excitation only during the evolution period, while the B₀ field has its low value of B₀ ^e, the required power level of the EPR excitation irradiation is minimised and that this is achieved without compromising the output image signal SNR which is determined by the higher value of the B₀ field during the other periods of the cycle.

In principle, the reduction of the B₀ field during the evolution period could be achieved by reducing the energising current of the colls 11 and 12, but in practice that would probably place unacceptable demands on the coll power supply and coil insulation due to the large inductance of the colls. It 1s therefore preferred to use the above described "field compensation technique" in which the colls 11 and 12 are maintained at constant current and the field varied by switching on and off only the much smaller colls 14. It 1s preferred that the colls 14 are switched off during the detection period rather than operated at reduced current, since then the B₀ field is provided only by the colls 11 and 12 during the detection period, when the greatest demands are placed on the spatial homogeneity and temporal stability of the magnetic field at the sample. The homogeneity of the magnetic field during the evolution period need only be good enough to irradiate the EPR line of Interest throughout the sample: in the conditions described above, the linewidth was more than 4MHz at an EPR frequency of 160 MHz so that a variation of B₀ ^e of more than ± 1% over the sample volume could be tolerated, while the calculε- d homogeneity of the coils 14 was better than ± 1000 pp over the sample volume. The disadvantage of field compensation is the inevitable interaction between the coils 11 and 14 caused by their close proximity: it was found that this gave rise to an instability of the magnetic field due to coils 11 and 12 when the current in the coils 14 was switched. The effect became more serious as the field strengths were increased, and 1t was for th∑ reason that the values of B_QP and m? were restricted to an upper limit of 0.01 T. In carrying out the method of the invention there may also be included the additional steps of obtaining an unenhanced NMR image signal by applying the magnetic field cycling NMR method as above described but without exciting EPR resonance, and then deriving image information representing the difference between the enhanced and unenhanced Image signals.

In a development of the invention, the method of the Invention may be used to obtain EPR spectral information from a sample. The EPR spectra may be used to distinguish between a mixture of paramagnetic species (or free radicals) in the sample. It is known that the concentration of molecular oxygen dissolved in a sample affects the observed enhancement E. However, E is also affected by the local concentration of the free radical under study, which may not be known. It is known that measurements of the superhyperfine structure of the EPR resonance allows the concentration of dissolved oxygen to be derived (P. D. Morse and H. M. Swartz, Magn. Reson. Med. 2_, 275 (1985)), and In principle this enables the oxygen concentration to be derived independently of the free radical concentration (or vice-versa). The method of achieving this relies on the dependence of the enhancement on the frequency of the EPR irradiation relative to the EPR line of interest. The larger the frequency (or magnetic field) offset from the centre of the EPR line, the lower 1s the enhancement. The sample is therefore examined several times by the method of the invention, and a series of images is collected, each image being collected with the same EPR irradiation frequency and the same values of B_QP and B_Q ^d but with a different value of B₀ ^e for each examination. A range of values of B₀ ^e is chosen so that the full width of the EPR line of interest is covered, the spacing between consecutive B₀ ^e values depending on the amount of spectral detail required. After the images have been produced from the raw data, a plot of enhancement E versus image number (or B₀ ^e) can be made for every pixel in the field of view. This plot is related to the EPR spectrum of the free radical under study. In experiments using this development of the method of the invention, the coils 11 and 12 were energised to provide a constant field of 0.01 T, giving an NMR frequency of 425 kHz. A double resonance coil assembly was used consisting of a solenoidal NMR coil (inside diameter 4 cm) inside a birdcage EPR resonator tuned to 288 MHz. The magnetic field offset during the EPR irradiation was software-selectable and was achieved by appropriate choice of the current through the coils 14, driven by a standard gradient amplifier. This arrangement allowed very small, reproducible changes in B₀ ^e to be made, so that the EPR line could be studied in detail. Spectral images were obtained using a sample of 2 M TEMPOL free radical. A saturation-recovery spin-warp pulse sequence was used with a TR of 1000 ms. An interleaved pulse sequence was used in which alternate NMR 90° pulses were preceded by an 800ms EPR irradiation. Thus pairs of images with and without EPR irradiation were obtained simultaneously. After processing, the pairs of images were combined subtractively to produce an image showing only those regions of the sample containing free radical. Three series of 20 32 x 32 Images of this kind were collected, each series covering one of the TEMPOL resonances centred on 0.0094 T, 0.010 T and 0.0118 T, respectively. Region-of-1nterest software then allowed an EPR spectrum to be obtained from any part of the field of view. The EPR spectral data were best visualised if the subtracted images (showing only the free radical) were used, since the image intensity fell to zero monotonically to the centre of the EPR line. When the non-subtracted images were used to obtain the EPR spectra, the image magnitude fell to zero and then increased to the signal level obtained without EPR irradiation. This was due to the fact that the enhancement factor E is negative when the EPR irradiation is near to the centre of the EPR Line but is positive and less than unity when the EPR irradiation is far from the centre of the line.

Claims

WE CLAIM:

1. A method of obtaining image information representing the distribution of paramagnetic material in solution in a solvent, comprising the steps of obtaining an NMR image signal of selected nuclei of the solvent by the magnetic field cycling NMR method in which each cycle includes a polarisation period followed by an evolution period followed in turn by a detection period, and only during the evolution period of each cycle, applying radio-frequency radiation at a frequency appropriate to the magnetic conditions than appertaining so as to excite EPR resonance in the paramagnetic material, thereby to enhance the NMR signal obtained from those selected nuclei which interact with electrons of the paramagnetic material excited by the radio-frequency radiation.

2. A method as claimed in Claim 1 and including the steps of maintaining the main (B₀) NMR magnetic field at a constant first value during the said polarisation period, then reducing that field to a second value, lower than said first value, and maintaining 1t constant at said second value during said evolution period, and then increasing that field to a third value, greater than said second value and maintaining it constant at said third value during said detection period.

3. A method as claimed in Claim 2, wherein said third value is less than said first value.

4. A method as claimed in Claim 2, wherein said third value is equal to said first value.

5. A method as claimed in Claim 1 and including the further steps of obtaining an unenhanced further NMR image signal of said selected nuclei by said magnetic field cycling NMR method carried out in the absence of excitation of EPR resonance in said paramagnetic material, and then deriving differential image information representing the difference between the enhanced NMR signal and said unenhanced further NMR signal.

6. A method as claimed in Claim 2 and including the further steps of obtaining an unenhanced further NMR image signal of said selected nuclei by said magnetic field cycling NMR method carried out in the absence of excitation of EPR resonance in said paramagnetic material, and then deriving differential image information representing the difference between the enhanced NMR signal and said unenhanced further NMR signal.

7. A method of obtaining image information representing the respective distributions of a plurality of paramagnetic materials in solution in a solvent, comprising the steps of repeatedly carrying out the method of Claim 6, on each repetition selecting different values as the said second value of the main (B₀) NMR magnetic field which is maintained during evolution periods of that repetition, thereby obtaining a series of differential image results together representing a spatial distribution of the EPR resonance response spectrum of one of said paramagnetic materials, and deriving therefrom information of the spatial distribution of the said one, and of another, of said paramagnetic materials.