Observation of axial magnetization of an object in a magnetic field

Abstract

A method of observing axial magnetization (Mz) in an object (O) located in a main magnetic field (B.sub.0) comprises the step of determining magnetic field intensity (B.sub.p) in at least one magnetic field probe (P) arranged in the neighborhood of the object. The magnetic field probe comprises a magnetic resonance (MR) active substance, means for pulsed MR excitation of the substance and means for receiving an MR signal generated by said substance.

Claims

1. A method of observing axial magnetization (Mz) in an object (O) located in a main magnetic field (B.sub.0), determining magnetic field intensity (B.sub.P) in at least one magnetic field probe (P) arranged in the neighborhood of the object, said magnetic field probe comprising a magnetic resonance (MR) active substance, means for pulsed MR excitation of said substance and means for receiving an MR signal generated by said substance, said step of determining magnetic field intensity (B.sub.P) comprising pulsed excitation of said substance with said means for pulsed MR excitation, said step of determining magnetic field intensity being carried out in a plurality of at least two magnetic field probes (P.sub.i, with i=1 to n, n≧2), thereby providing respective magnetic field intensities (B.sub.Pi, with i=1 to n, n≧2), said method further comprising the step of subtracting a background magnetic field (B.sub.B) from each one of said respective magnetic field intensities Bpi, said step of determining magnetic field intensity being repeated at a predetermined sampling rate, said sampling rate being up to 100 Hz.

2. The method according to claim 1, wherein said background magnetic field (B.sub.B) is modeled as a linear combination of preselected basis functions ƒ.sub.1(r).

3. The method according to claim 2, wherein said pulsed excitation induces nuclear magnetic resonance of said substance.

4. The method according to claim 2, further comprising the step of obtaining from said magnetic field intensity (B.sub.P) an observable (Az) that is proportional to said axial magnetization (Mz).

5. The method according to claim 2, further comprising the step of manipulating nuclear magnetization of an MR active nuclear species in said object by applying at least one radiofrequency field and optionally at least one gradient field before or concomitantly with said observing of axial magnetization (Mz).

6. The method according to claim 1, wherein said pulsed excitation induces nuclear magnetic resonance of said substance.

7. The method according to claim 6, further comprising the step of obtaining from said magnetic field intensity (B.sub.P) an observable (Az) that is proportional to said axial magnetization (Mz).

8. The method according to claim 6, further comprising the step of manipulating nuclear magnetization of an MR active nuclear species in said object by applying at least one radiofrequency field and optionally at least one gradient field before or concomitantly with said observing of axial magnetization (Mz).

9. The method according to claim 6, further comprising the step of manipulating nuclear magnetization of an MR active nuclear species in said object by applying at least one radiofrequency field and optionally at least one gradient field before or concomitantly with said observing of axial magnetization (Mz).

10. The method according to claim 1, further comprising the step of obtaining from said magnetic field intensity (B.sub.P) an observable (Az) that is proportional to said axial magnetization (Mz).

11. The method according to claim 10, further comprising the step of manipulating nuclear magnetization of an MR active nuclear species in said object by applying at least one radiofrequency field and optionally at least one gradient field before or concomitantly with said observing of axial magnetization (Mz).

12. The method according to claim 1, further comprising the step of manipulating nuclear magnetization of an MR active nuclear species in said object by applying at least one radiofrequency field and optionally at least one gradient field before or concomitantly with said observing of axial magnetization (Mz).

13. The method according to claim 12, further comprising the step of determining a nuclear relaxation property of said MR active nuclear species from the temporal behavior of said observed axial magnetization (Mz).

14. The method according to claim 1, wherein the object of interest is a material sample, a water sample, a sample of body liquid, a cell culture, or a plant or part thereof.

15. The method according to claim 1, wherein the object of interest is a live human or animal, wherein said at least one magnetic field probe is mounted on a chest region, for observing field fluctuations caused by the beating heart.

16. The method according to claim 1, wherein the object of interest is a live human or animal, wherein said observing of axial magnetization (Mz) is used to control an MRI procedure carried out in said main magnetic field.

17. The method according to claim 1, wherein the object of interest is a live human or animal, wherein said observing of axial magnetization (Mz) is used to monitor status and/or compliance of said human or animal.

18. The method according to claim 1 wherein the object of interest is a live human or animal, wherein said observing of axial magnetization (Mz) is used to to support signal processing anellor image reconstruction.

19. The method according to claim 1, wherein the object of interest is a live human or animal, wherein said at least one magnetic field probe is mounted on a head region, wherein said observing of axial magnetization (Mz) is used to observe field fluctuations caused by physiological processes in the brain.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The above mentioned and other features and objects of this invention and the manner of achieving them will become more apparent and this invention itself will be better understood by reference to the following description of various embodiments of this invention taken in conjunction with the accompanying drawings, wherein are shown:

(2) FIG. 1 a) a basic arrangement for the observation of a sample magnetization by means of one NMR field probe;

(3) b) a further arrangement for the observation of a sample magnetization by means of four NMR field probes;

(4) FIG. 2 the magnetic field (expressed in Hz) as a function of time, measured with an NMR field probe at three positions on a person's chest along the sternum.

DETAILED DESCRIPTION OF THE INVENTION

(5) To carry out the methods of this invention, the object to be examined is placed in a background field, which magnetizes the object according to its electronic and nuclear susceptibility distributions. Preferably, the background field should be approximately uniform to magnetize the object evenly. Highly uniform magnetic fields as commonly used for NMR or MRI are suitable for this purpose. However, the invention can also be carried out in substantially less uniform fields.

(6) One or multiple MR-based field probes are placed in the vicinity of the object. A single probe is preferably placed close to that part of the object that is of most interest. For observing cardiac dynamics, for instance, it should be placed on the chest wall close to the heart. A set of multiple field probes disposed on or near the surface of the object is used to achieve spatial resolution of the magnetic field generated by Mz beneath the surface. When examining a largely uniform sample such as a vial filled with a sample material, multiple field probes can be arranged to surround the sample.

(7) The T1 and T2 relaxation times of the MR probes are chosen such as to support the desired rate of re-excitation. In a preferred mode of operation, the probes are re-excited for each field measurement, T2 is set as long as possible to still suppress significant echo formation by the interaction of successive excitation pulses, and T1 is chosen similar to T2.

(8) In the arrangement shown in FIG. 1 a), there is an object (O) located in a main magnetic field (Bo). Typically this will be a person or animal, or a body part thereof, or any other object of interest. In this basic set-up there is just one single magnetic field probe (P) arranged in the neighborhood of the object. In this example the magnetic field probe comprises a magnetic resonance (MR) active substance 2 enclosed inside a small tube and means such as a surrounding solenoid 4 serving for pulsed MR excitation of the MR active substance and also for receiving an MR signal generated by the MR active substance. It is understood that the excitation and receiving means comprise further components not shown in these drawings, particularly electronic components for generating the RF excitation pulse and other electronic components for receiving and processing the probe signal. It is also understood that a design with a single solenoid is merely one of many possible configurations.

(9) As may be appreciated from FIG. 1a, the magnetic field probe P allows determination of the magnetic field intensity (B.sub.P) at the probe location. By selecting an appropriate probe design with a fast response time, the information gained from the probe provides a rather accurate determination of the time dependent function B.sub.P (t), which in turn provides information about the axial magnetization (Mz) in the object of interest.

(10) FIG. 1b shows an improved arrangement with four magnetic field probes P1, P2, P3, P4 arranged around the object of interest. By using more than one probe located at different places and at different distances from the object of interest, it is possible to better discriminate the requested Mz from any other spatiotemporal “background” effects.

(11) To calculate the local field intensity from an interval of MR probe signal, first the signal phase (in radians) is extracted. The signal phase is uniquely defined only up to multiples of 2π. This ambiguity is removed by standard phase unwrapping along the temporal dimension, i.e., by adding a suitable integer multiple of 2π to each phase value such that the resulting phase time course is continuous. The average angular frequency of the signal during the subject interval is then calculated by weighted linear regression of the unwrapped phase time course, using the magnitude of the probe signal for weighting. The corresponding field value for said interval is obtained by dividing the angular frequency by 2π and by the gyromagnetic ratio of the probe substance. Repeating this process for successive signal intervals yields a time series of field values.

(12) Such a field time series represents the sum of the background field and its potential fluctuations, a small constant field shift induced by the magnetization of the field probe itself, and the desired field contribution from Mz in the object. The constant contributions of the background field and the probe-specific shift can easily be subtracted based on an initial reference field measurement with the same probe. However, fluctuations of the background field and fluctuations of Mz cannot straightforwardly be distinguished for single probes or sets of probes mounted on heterogeneous objects such as a human body.

(13) However, in the case of a substantially uniform sample, e.g., a vial filled with a sample material, in a substantially uniform background field, a set of multiple field probes disposed around the sample can serve to eliminate background field fluctuations of low spatial order. To this end, it is necessary to know the probe positions r.sub.P either from reference measurements, e.g., based on frequency measurements in the presence of well-defined gradient fields, or by mounting them precisely in the first place. For each set of field values obtained in a given interval, one from each probe, processing then starts again by individually subtracting a probe-specific reference field value measured once initially and simultaneously with all probes. The resulting field values for said interval are conveniently assembled in a column vector b having one entry per probe. To remove the effects of background fluctuations from b, the underlying fields can be modeled as a linear combination of known basis functions ƒ.sub.1(r). For magnetic fields in a source-free volume, spherical harmonics are a favorable basis. Evaluating these basis functions at the probe positions r.sub.P yields the matrix F having one row per probe and one column per basis function. Removal of background field fluctuations is then achieved by removing all components from b that lie in the space spanned by the columns of F. This is achieved by calculating b′=(Id−FF.sup.+)b where the superscript+denotes the Moore-Penrose pseudoinverse. The resulting field values listed in b′ now all scale with the strength of Mz in the sample. For this approach to work, the number of probes must be larger than that of basis functions and they must be placed such that the matrix F has full rank.

Example: Observation of Cardiovascular Dynamics by Field Recording with an NMR Probe

(14) Targeting a temporal resolution of 10 ms, an NMR probe was built from a 2.2 mm borosilicate capillary filled with water and doped with GdCl.sub.3 such as to obtain fast transverse relaxation with T.sub.2=3 ms. For RF transmission and reception, the capillary was placed in a tightly wound solenoid coil made from PTFE-coated silver wire. The coil was tuned to 297.8 MHz, matched, and connected to custom-built transmit/receive circuitry including pre-amplification. Via coaxial cable, the preamplified signal was fed into a laboratory spectrometer (National Instruments) for demodulation and recording at a bandwidth of 1 MHz. The spectrometer was configured for continuous signal reception. The probe was excited via a custom-built transmit chain, consisting of a pulse generator, a modulation stage and a power amplifier.

(15) For field measurements, the probe was excited every 10 ms with 90° block pulses of 10 μs each and its signal was received continuously, resulting in interruptions only by the excitation pulses and a few additional us of T/R switching and filter delay. The signal time course was then segmented into individual FIDs of just under 10 ms duration. The phase time courses of the FIDs were calculated, unwrapped, and subject to linear regression to give one frequency measurement per 10 ms interval.

(16) Measurements were performed on a healthy volunteer in the bore of a Philips 7T Achieva whole-body MR system (Philips Healthcare, Cleveland, USA). The field probe was placed on the volunteer's chest, starting from the center of the sternum and gradually shifting it towards the head in steps of 2 cm. At each position, a field measurement was performed during a breathhold of 8 seconds.

(17) FIG. 2 shows resulting time courses of frequency variation for the first three probe positions. They reveal substantial, highly periodic field variations at the cardiac frequency, which are attributed to motion of the heart and blood flow in the heart and the neighboring vasculature. As should be expected, the shape of these curves depends on the position of observation, indicating that the field fluctuations vary significantly across space and thus contain information about the position and geometry of the underlying anatomy.

(18) The exact interpretation of the curves is intriguing and remains to be pursued. The most immediate question is arguably which type of motion causes the two distinct field peaks that are observed in each cycle at all three positions. They may reflect myocardial motion but may also be due to particular phases of blood flow in the heart and/or ascending aorta. Further studies will need to establish which aspects of cardiovascular dynamics are actually observed and how much more evidence can be gained by increasing the temporal and spatial resolution of the measurement. Increasing the latter is expected to be fairly straightforward by using an array of field probes. Increasing the temporal resolution will, among others, clarify the effective bandwidth of the field dynamics. The sharp negative peaks in FIG. 1 c) consist of single field samples, indicating that the underlying fluctuations last for less than 10 ms, thus suggesting a bandwidth of somewhat more than 100 Hz. If this observation can be confirmed and perhaps even higher-bandwidth dynamics can be observed, NMR-probe measurements may offer an alternative to ECG recording in the magnet.