Magnetometer arrays operating in zero field and related methods for calibrating couplings between magnetometers

Abstract

The present invention relates to a method for determining the coupling between magnetometers in an array of N magnetometers (e.g., by means of an optically pumped magnetometer), wherein each magnetometer includes a field cancellation system that can be enabled to operate the magnetometer in a zero field. The method includes a first phase (P1), in which the N magnetometers are divided into N - 1 magnetometers with the field cancellation system deactivated and one measuring magnetometer with the field cancellation system activated. The first phase includes: - generating (GENj) by these magnetometers a plurality of reference magnetic fields with known amplitudes and different directions, - measuring (MESi) by the measuring magnetometer the ambient magnetic field on a plurality of measurement axes, - determining (CALCij) the coupling coefficient between the measuring magnetometer and each of the N magnetometers based on the measured values and these known amplitudes.

Description

Field of the Invention

[0001] The field of the invention is that of biomagnetic imaging, and more particularly the invention relates to an array of vector magnetometers especially for magnetocardiography or magnetoencephalography. Background Art

[0002] The imaging of the magnetic fields generated by different organs of the human body (in particular the heart and the brain) uses a matrix of magnetic sensors (magnetometers) to identify information that is relevant both from a medical research perspective and from a diagnostic perspective for different pathologies. Thus, electrocorticogram imaging allows the study of rapid brain dynamics in a non-invasive manner. It is also used for the pre-operative diagnosis of epilepsy and can more accurately localize the epileptogenic zone compared to electroencephalography.

[0003] In order to reconstruct the magnetic field sources (such as the currents circulating in the heart or the brain), it is necessary to obtain the magnetic field measurements along one or different measurement axes and the positions of the sensors that record these measurements.

[0004] Currently, magnetoencephalography often uses hundreds of SQUID magnetic sensors, which have a very low intrinsic noise level but require cryogenic cooling to operate. In addition, these imaging techniques are generally implemented within a magnetically shielded enclosure to eliminate external magnetic fields that may interfere with the measurements. These enclosures must be large enough to accommodate the cryostat, which places significant limitations on the cost and the architecture of the building containing these devices, and these limitations are not conducive to the popularization of these imaging techniques.

[0005] Alternatively, optically pumped magnetometers with a similar intrinsic noise level but that do not require cooling can be used. Thus, the sensors can be placed closer to the patient's skin, which can increase the amplitude of the signal and the spatial resolution of the measurements. It is also possible to reduce the magnetic shielding, which allows these techniques to be considered for more widespread use in hospitals.

[0006] In addition, optically pumped magnetometers allow the position of the sensors to be adapted to the body surface of each specific patient. This is an advantage, but the position of the sensors must be recalibrated each time the array is adjusted according to the morphology of a new patient. Therefore, it is important to be able to effectively calibrate the positions of the different magnetometers relative to each other, as well as the relative position of the array relative to a reference element positioned on the patient.

[0007] Many parameters for the positioning and calibration of optically pumped magnetometers (especially gain and linearity) can be stabilized (without the need for continuous measurement) by operating the magnetometers in a closed loop, which consists of generating a field called the compensation field by reacting on the coils of each magnetometer so that each magnetometer operates in a zero total magnetic field (ambient field + compensation field) along its different measurement axes. This closed-loop operating mode has been used for magnetocardiography and magnetoencephalography measurements as reported in the following publication: E. Labyt et al., "Magnetoencephalography With Optically Pumped 4 He Magnetometers at Ambient Temperature [Magnetoencephalography with Optically Pumped Helium-4 Magnetometers at Ambient Temperature]", IEEE Transactions on Medical Imaging, vol. 38, n°1, p. 90-98, jan. 2019 [IEEE Transactions on Medical Imaging, Volume 38, Number 1, Pages 90 - 98, January 2019].

[0008] However, only two sensors were implemented in that publication, which were far enough apart from each other so as not to be significantly affected by the compensation field of the other sensor. However, for useful medical measurements, dozens or even hundreds of such sensors need to be implemented, which makes it necessary for them to be closer to each other. The resulting drawback is the coupling between magnetometers placed close to each other, which are affected by the compensation fields of other magnetometers. Therefore, it is no longer possible to obtain the value of the magnetic field that pre-existed during the operation of the magnetometers by simply reading the output of the sensors.

[0009] However, there are procedures that allow the recovery of these field values by reading the output of the sensors and by reading a matrix that transforms the coupling between different compensation coils and the magnetometers of the array. Thus, patent application EP 3 299 831 A1 describes a way to correct the coupling in an array of magnetometers operating in a closed loop, as well as a technique for determining the coupling matrix, which can have two variants: one that measures the coupling with all magnetometers operating in an open loop and another that measures the coupling with different magnetometers operating successively in a closed loop.

[0010] The first of these methods (open-loop calibration) yields a true coupling coefficient only when operating in an environment where the residual field is nearly zero. However, the inventors have been able to experimentally observe that when there is a field of a few nanoteslas, a significant (up to 17%) coupling is measured, while the actual coupling is less than 1%. This artifact seems to stem from second-order effects that make parametric resonance and Hanle effect magnetometers sensitive not only to the field along their nominal measurement axis (e.g., B along the z-axis z ) but also have cross terms such as B x , B y .

[0011] In addition, for the case of parametric resonance magnetometers, as described in the aforementioned patent, the coupling between the coil of the sensor and the sensitive element of another sensor close to the first sensor also brings additional drawbacks. In such a magnetometer configuration, an RF field is applied along one or more axes of the sensor, which define the geometric axes for measuring different components of the magnetic field. Thus, the coupling between two magnetometers with different axis directions causes a change in their respective measurement axes, which has a very negative impact on the reconstruction accuracy of the magnetic field source. Summary of the Invention

[0012] The present invention aims to propose a method for determining the coupling between different magnetometers in an array of optically pumped magnetometers (e.g., Hanle effect or parametric resonance magnetometers), which overcomes the above-mentioned drawbacks.

[0013] To this end, the present invention relates to a method for determining the coupling between magnetometers in an array of N magnetometers, where each magnetometer includes a field cancellation system that can be enabled to operate the magnetometer in a zero field. The method includes a first stage in which the N magnetometers are divided into N - 1 magnetometers with the field cancellation system deactivated and one measuring magnetometer with the field cancellation system enabled. The first stage includes the following steps:

[0014] ○ Generating, by each of the N magnetometers (i.e., the measuring magnetometer and the N - 1 magnetometers with the field cancellation system deactivated), a plurality of reference magnetic fields with known amplitudes and different directions,

[0015] ○ Measuring, by the measuring magnetometer, the ambient magnetic field on a plurality of measurement axes;

[0016] ○ Determining the coupling coefficient between the measuring magnetometer and each of the N magnetometers (i.e., the measuring magnetometer and each of the N - 1 magnetometers with the field cancellation system deactivated), said determination including:

[0017] ■ Detecting the contribution of the reference magnetic field to the ambient magnetic field measurement on one of the measurement axes;

[0018] ■Calculate the ratio between the magnitude of the contribution of the reference magnetic field and the known magnitude of the reference magnetic field.

[0019] Some preferred but non - limiting aspects of the method are as follows:

[0020] Iterate the steps of the first stage N times by using a new magnetometer in the array as the measuring magnetometer in each iteration;

[0021] Each reference magnetic field carries its specific item of information, and detecting the contribution of the reference magnetic field includes identifying the information specific to the reference magnetic field;

[0022] The information specific to the reference magnetic field is a characteristic frequency, such as a high fundamental frequency of a given power of a number;

[0023] The magnetometers are parametric resonance magnetometers, where each parametric resonance magnetometer includes an excitation system that can be enabled to induce a parametric resonance excitation radio - frequency field, and in the first stage, the excitation systems of the N magnetometers are enabled;

[0024] The method includes a second stage that is the same as the first stage (P1), except that the excitation systems of N - 1 magnetometers other than the measuring magnetometer are deactivated;

[0025] The second stage includes N iterations, with each iteration using a different measuring magnetometer among the N magnetometers. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Other aspects, objects, advantages, and features of the present invention will become more apparent by reading the following detailed description of the preferred embodiments of the present invention given by way of non - limiting examples and with reference to the accompanying drawings. In the accompanying

[0027] FIGURES:

[0028] Figure 1 is a diagram of a magnetometer belonging to a magnetometer array according to the present invention;

[0029] Figure 2 is a diagram showing different steps of the calibration method according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0030] The present invention relates to a magnetic field measuring device, and more particularly to a device equipped with N (a natural integer of at least 2) vector magnetometers arranged in an array. The device has a particular application in the medical field of biomagnetic field imaging (especially in magnetoencephalography or magnetocardiography).

[0031] The present invention more particularly relates to a device configured to be able to implement the following method for calibrating a magnetometer. This method can estimate the influence of an array of magnetometers by measuring the interference caused by the operation of one of the magnetometers on the measurements performed by another magnetometer.

[0032] In order to apply this method and, more generally, to be able to successfully operate an array of magnetometers in a closed-loop manner, preferably, the magnetometers of the array are capable of measuring the three components of the magnetic field. In fact, in the contrary case, closed-loop operation cannot guarantee the absence of a residual field in any of the magnetometers of the array. This is therefore a necessary condition for the coupling of the measurements to be free of the above-mentioned artifacts.

[0033] Without limiting the present invention, the magnetometers of the array are preferably optically pumped magnetometers, such as Hanle effect magnetometers or parametric resonance magnetometers.

[0034] Therefore, the following description is given by way of example of an optically pumped magnetometer. Referring Figure 1 , each magnetometer 20 of the array includes a cell 1 filled with an atomic gas (such as helium-4 or an alkaline gas), which is affected by the ambient magnetic field, the projections of which on three mutually perpendicular coordinate axes x, y, z define three components. The ambient magnetic field is thus decomposed into three components Bx, By, and Bz respectively according to one of the measurement axes x, y, and z of the magnetometer.

[0035] The cell 1 is irradiated by optical pumping sources 2, 3, which are arranged to emit a beam of light, such as a laser beam, in the direction of the cell 1, the beam being tuned to the pumping wavelength (this beam is therefore also designated as the pumping beam). The pumping wavelength is wedged on an atomic transition line, such as the line D0 at 1083 nm in the case of helium-4. The beam can be emitted by a laser source 2 and linearly polarized by a linear polarizer 3, which is inserted between the laser source 2 and the cell 1 or directly integrated into the laser source 2. The beam propagates in a propagation direction coinciding with the x-axis and is linearly polarized along the z-axis.

[0036] In the case where the sensing element is helium-4, the magnetometer 20 further includes a high-frequency (HF) discharge system, which includes a generator HF 4 and an overvoltage coil 5 to put the atoms of the atomic gas in an excited state, in which state, when these atoms are irradiated by a beam that is normally in the metastable state 2 3 S1, they are able to undergo atomic transitions.

[0037] When the magnetometer 20 is a parametric resonance magnetometer, it further includes an excitation system that can be enabled to sense the parametric resonance excitation radio frequency field. The parametric resonance excitation system includes a radio frequency generator 8 that provides orthogonal axes around the unit for a Helmholtz coil 7 to generate a parametric resonance excitation magnetic field, also known as an excitation radio frequency field. The excitation circuit more specifically generates a radio frequency magnetic field having two components orthogonal to the polarization direction and each component oscillating at its own oscillation frequency. The two components are the component B ω cosωt oscillating along the x-axis at a pulsation ω (e.g., ω = 2π·3000 kHz) and the component B Ω cosΩt oscillating along the y-axis at a pulsation Ω (e.g., Ω = 2π·16 kHz). These components result in resonances at each of the oscillation frequencies Ω / 2π, ω / 2π and interharmonics at the oscillation frequency (ω±Ω) / 2π, and these resonances are respectively associated with the values of the ambient field in the x, y, and z directions.

[0038] The magnetometer 20 further includes: a photodetector 6 arranged to receive the light beam that has passed through the unit; and a parametric resonance detection circuit 9 configured to perform synchronous detection at the harmonics of each oscillation frequency of the electrical signal transmitted by the photodetector and at the interharmonics of the oscillation frequency of the electrical signal transmitted by the photodetector when the magnetometer 20 is a parametric resonance magnetometer. The device 9 includes three detection channels: a first channel Vx for detecting the signal at Ω / 2π (x-axis), a second channel Vy for detecting the signal at ω / 2π (y-axis), and a third channel Vz for detecting the signal at Ω±ω / 2π (z-axis). The signals on each of the first channel Vx and the second channel Vy are first amplified and then filtered through a band-pass filter corresponding to an appropriate center frequency (i.e., corresponding to the center frequency of the applied RF field). Then the obtained signals are multiplied by a reference signal and processed through synchronous detectors DSx, DSy. The third channel Vz continuously employs two synchronous detections, one through the detector DSy of the second channel Vy at ω / 2π and the other through the synchronous detector DSz at Ω / 2π.

[0039] The magnetometer 20 further includes a field cancellation system that can be enabled to operate the magnetometer in a zero field. This system can take the form of the servo control circuit 10 of a closed-loop magnetometer. Such a servo control circuit includes three servo control channels Wx, Wy, and Wz, which are respectively coupled to the outputs of the corresponding detection channels Vx, Vy, and Vz. The servo control channels Wx, Wy, and Wz respectively use the outputs of the corresponding detection channels as error signals to continuously readjust the compensation field. The servo control channels Wx, Wy, and Wz respectively include integrators Ix, Iy, and Iz configured to transmit compensation signals and current generators GCx, GCy, and GCz driven by the compensation signals to inject current into one of the Helmholtz coils 7 in order to generate compensation magnetic fields BCx, BCy, and BCz opposite to the components Bx, By, and Bz of the ambient field. By measuring the current flowing in the coil 7, the fields that must be applied to cancel the different components of the ambient field and thus have the values of these different components can be deduced.

[0040] The magnetometer 20 is also provided with a circuit for generating a reference magnetic field, and the measuring device (i.e., the magnetometer array) itself includes a computer configured to drive the magnetometers of the array and determine the coupling coefficients between the magnetometers according to the method to be described below. The reference magnetic field generation circuit can include a reference signal generator 11, which is connected in parallel with the generator 8 for connecting the coil to the parametric resonance excitation system and the circuit of the servo control circuit 10 of the closed-loop magnetometer to each Helmholtz coil 7. This connection of the reference signal generator 11 to each Helmholtz coil 7 can be achieved through a bias resistor, the value of which is higher than the impedance of the coil at low frequencies (<kHz). Then, the sum of the currents on the coil allows maintaining the dynamics specific to the magnetometer unchanged while generating the required reference field.

[0041] Due to the magnetometer array, in the normal use of the array, the magnetic field measured at the magnetometer integrates the contributions of all other magnetometers. Therefore, the magnetic field B i (actually by the coil of the magnetometer M i ) generated Gi induces a field B at the magnetometer M j (actually on each measurement axis of the magnetometer M j ) according to the coupling coefficient Mj .

[0042] An attempt is made to determine the coupling matrix consisting of different coupling coefficients during the calibration operation before the normal use of the magnetometer array. The size of this matrix is 3N x 3N, where N corresponds to the number of magnetometers in the array, and the number 3 represents the three axes of the magnetometer. Therefore, B Mj = ∑ i Cij ·B Gi where the indices i and j range from 1 to 3N, and each component of the field generated by the magnetometer can be seen on each measurement axis of the magnetometer (thus obtaining three measurement values B Mj which respectively integrate the contributions of the three fields B Gi in each of the fields).

[0043] In the measurement phase after this calibration phase, by using the inversion of the coupling matrix, it is possible for each magnetometer to return to the actual magnetic field as it would be measured in the absence of the array.

[0044] The following refers to Figure 2 the different steps of this calibration operation. This operation uses the coil 7 of the magnetometer to generate a magnetic field of known amplitude by means of the reference signal generator 11. The generated field is successively measured by one of the magnetometers in the array to establish the coupling matrix and its 3N x 3N coefficients.

[0045] This calibration operation includes a first phase P1 which includes N iterations, each iteration on a different measuring magnetometer among the N magnetometers.

[0046] In each iteration of the first phase P1, the N magnetometers are divided into N - 1 magnetometers with the field cancellation system deactivated and one measuring magnetometer with the field cancellation system activated. When the magnetometers are of the parametric resonance type, their parametric resonance excitation systems are activated in the first phase. Thus, the measuring magnetometer operates in a zero field and performs a field measurement (its field cancellation system is activated), while all the other magnetometers do not perform measurements and thus do not compensate for the field at their positions (their field cancellation systems are deactivated), and their role is limited to generating a reference field on their coils and, where appropriate, generating the radio frequency corresponding to their normal operation (their parametric resonance excitation systems are activated).

[0047] This way of working with a closed - loop measuring magnetometer ensures that the characterization of the coupling is made using this measuring magnetometer operating in a zero field along its three measurement axes, which avoids the appearance of a large number of artifacts caused by inter - axis effects that are detrimental to the characterization of the coupling. In addition, the fact that the parametric resonance excitation circuit is activated during this characterization of the coupling allows, when the magnetometers are of the parametric resonance type, to measure these couplings by taking into account the misalignment of the measurement axes that may be caused by the coupling between the radio frequencies of different magnetometers. These misalignments also exist during the operation of the sensor array and thus need to be taken into account when determining the coupling matrix.

[0048] Each iteration of the first phase P1 includes the following steps:

[0049] ○ Each of the N magnetometers generates GENj a plurality of reference magnetic fields with known amplitudes and different directions.

[0050] ○ The measuring magnetometer measures MESi the ambient magnetic field on a plurality of measurement axes.

[0051] ○ Based on the measured values of the ambient field and the known amplitudes of the reference fields, determine CALCij the coupling coefficient between the measuring magnetometer and the N magnetometers.

[0052] During the generation step GENj, the circuit for generating the reference magnetic field in each of the N magnetometers can thus be enabled, such that each of the N magnetometers generates a reference field with a known amplitude on each of these three axes (three different directions). And, during the measurement step MESi, the measuring magnetometer measures the ambient magnetic field on each of its three measurement axes.

[0053] It is possible that the amplitude of the reference field is too high and causes adverse effects, such as eddy currents in metal components located near the array. In this case, the amplitude can be reduced, for example, by one order of magnitude relative to the amplitudes of other reference fields, and this reduction is then taken into account when calculating the coupling coefficient.

[0054] Determining CALCij the coupling coefficient includes detecting the contribution of each reference magnetic field to the ambient magnetic field measurement on one of the measurement axes. For each reference magnetic field, this detection is followed by calculating the coupling coefficient, which is in the form of the ratio of the amplitude of the contribution of the reference magnetic field to the known amplitude of the reference magnetic field. For each measurement axis, the measured coupling coefficient can be normalized by the self - coupling of that axis, i.e., the ratio of the amplitude of the contribution of the reference field generated on that axis to the known amplitude of the reference field on that axis.

[0055] Given that the coupling coefficients are symmetric (the coupling of i with j is the same as the coupling of j with i), the coupling coefficients calculated in both directions can be used to identify possible problems in the calibration method, or even only half of these coefficients can be calculated.

[0056] During the generation step GENij, the reference magnetic field generation circuit of each magnetometer can generate a reference field, each reference field carrying its specific information item (i.e., the axis - specific information item of that magnetometer), and this information can be identified in the ambient magnetic field measurement performed on one of the axes of the measuring magnetometer in order to isolate the contribution of the reference magnetic field carrying this information.

[0057] The information specific to the reference magnetic field can be a characteristic frequency. The reference signal generated by the signal generator 11 can, for example, consist of sine curves at frequencies specific to each axis of each magnetometer. For example, these frequencies can be selected at the harmonics corresponding to the high fundamental frequency of a given power, i.e., where f is the fundamental frequency and p is the given number. Thus, if the fundamental frequency is 1 Hz and if the given number is 2, the characteristic frequencies are the frequencies of 1, 2, 4, 8, 16, 32... Hz. The advantage of this choice of characteristic frequencies is that the presence of any harmonic that does not correspond to a power of the given number constitutes a good means of diagnosing unwanted inter-axis effects. For example, the first magnetometer can generate a reference field B1sin(2πt) on its x-axis, a reference field B1 sin(2π2t) on its y-axis, and a reference field B1 sin(2π4t) on its z-axis, while the second magnetometer generates a reference field B1sin(2π8t) on its x-axis, a reference field B1sin(2π16t) on its y-axis, and a reference field B1 sin(2π32t) on its z-axis, and so on. Detecting the contribution of the reference magnetic field to the ambient magnetic field measurement on one of the measurement axes of the measuring magnetometer can then include synchronous detection at the characteristic frequency of the reference field.

[0058] Alternatively, this information specific to the axis of the magnetometer (the reference field generated by the magnetometer) can correspond to the durations of the high and low phases of a reference signal in the form of time slots. In another variant, the reference signal forms an orthogonal basis, such as the orthogonal basis used in radio transmission technology by means of spectral spreading, where each magnetometer marks its transmission on one of its axes by a characteristic emission sequence in a different frequency band, and these frequency bands are included in the signal frequency band measurable by the array of magnetometers.

[0059] The measurement step MESi is performed successively, for example, on each axis of the measuring magnetometer. For example, first, a measurement is made along the first axis of the measuring magnetometer for a long enough time to satisfactorily detect the contribution of each reference magnetic field to the ambient magnetic field measurement on this first axis, for example, to distinguish reference fields with similar characteristic frequencies. Thus, when the fundamental frequency is 1 Hz (as in the above example), a time of 10 seconds has proven to be largely sufficient to avoid any overlap of the lines. Then, a similar procedure is performed successively on the second and third axes of the measuring magnetometer.

[0060] At the end of one iteration of the first stage, as Figure 2 shown in the box CHANi in, a new measuring magnetometer is selected among the N magnetometers as a new iteration of the first stage. Thus, the foregoing operations are repeated so as to be performed N times in total, each time using a different measuring magnetometer among the N magnetometers.

[0061] ReferenceFigure 2 , the calibration method according to the present invention may include a second stage P2, which also includes N iterations, as shown in block CHANi, and each iteration uses a different measuring magnetometer among the N magnetometers. The second stage may be implemented before or after the first stage. Alternatively, the iterations of the first stage and the second stage may be mixed, for example, by performing the iterations of the second stage after the iterations of the first stage using the same measuring magnetometer.

[0062] As previously known, enabling parametric resonance excitation RF in the first stage allows taking into account the misalignment of the measurement axis that may be caused by the coupling between the RFs of different magnetometers. However, these couplings are not conducive to estimating the orientation and position of different magnetometers in the array by triangulation or other more refined methods (such as the least squares estimation that takes into account the theoretical geometric configuration of the magnetic fields generated by each coil of each magnetometer).

[0063] For this reason, in the case of parametric resonance magnetometers, the method according to the present invention may include a second stage P2 that is the same as the first stage P1, except that the parametric resonance excitation systems of the N - 1 magnetometers other than the measuring magnetometer are deactivated during each iteration. Therefore, in the iterations of the second stage, the N magnetometers are divided into N - 1 magnetometers with the excitation system and the field cancellation system deactivated and one measuring magnetometer with the excitation system and the field cancellation system enabled. Refer to Figure 2 , this second stage includes the following steps:

[0064] ○ Generating, by each of the N magnetometers, a plurality of reference magnetic fields GENSEj with known amplitudes and different directions,

[0065] ○ Measuring, by the measuring magnetometer, the ambient magnetic field MESi on a plurality of measurement axes;

[0066] ○ Determining, based on the measured values of the ambient field and the known amplitudes of the reference fields, the coupling coefficients CALCij* between the measuring magnetometer and the N magnetometers.

[0067] The coupling matrix obtained at the end of the second stage P2 does not take into account the misalignment of the axis caused by the coupling between the RF of the magnetometer other than the measuring magnetometer and this RF. The coupling matrix determined in this second stage is only used to calibrate the positions and orientations of different magnetometers. Therefore, the nine coupling coefficients between two magnetometers can be used to establish the three angles and three distances between these two magnetometers, as described in, for example, patent EP 3 343 240 B1. The comparison between the matrices determined by the first stage and the second stage respectively also allows measuring the misalignment of the axis caused by RF coupling.

[0068] The invention is not limited to the method described above, but extends, as indicated previously, to a magnetic field measurement device configured to allow the implementation of the method. The device includes a computer and an array of N magnetometers, where each magnetometer includes a field cancellation system that can be enabled to operate the magnetometer in a zero field. The N magnetometers are particularly capable of performing the following steps while being divided into N - 1 magnetometers with the field cancellation system deactivated and one measuring magnetometer with the field cancellation system enabled:

[0069] ○ Generating (GENj) by each of the N magnetometers a plurality of reference magnetic fields having known amplitudes and different directions,

[0070] ○ Measuring (MESi) the ambient magnetic field by the measuring magnetometer on a plurality of measurement axes.

[0071] The computer, for its part, is specifically configured to detect the contribution of the reference magnetic field to the ambient magnetic field measurement on one of the measurement axes and to determine the coupling coefficient between the measuring magnetometer and one of the magnetometers by calculating the ratio between the amplitude of the contribution of the reference magnetic field and the known amplitude of the reference magnetic field.

Claims

1. A method for determining the coupling between magnetometers (20) in an array of N magnetometers, wherein, Each magnetometer includes a field cancellation system (10) that can be enabled to operate the magnetometer in a zero field, characterized in that the method includes a first stage (P1), in which the N magnetometers are divided into N - 1 magnetometers with the field cancellation system disabled and one measuring magnetometer with the field cancellation system enabled. The first stage includes the following steps: o Generating (GENj) by the measuring magnetometer and each of the N - 1 magnetometers with the field cancellation system disabled a plurality of reference magnetic fields with known amplitudes and different directions; o Measuring (MESi) the ambient magnetic field by the measuring magnetometer on a plurality of measurement axes; o Determining (CALCij) the coupling coefficient between the measuring magnetometer and each of the N - 1 magnetometers with the field cancellation system disabled and the measuring magnetometer, the determination including: ■ Detecting the contribution of the reference magnetic field to the ambient magnetic field measurement on one of the measurement axes; ■ Calculating the ratio between the amplitude of the contribution of the reference magnetic field and the known amplitude of the reference magnetic field.

2. The method according to claim 1, wherein The steps of the first stage are iterated N times by using a new magnetometer in the array as the measuring magnetometer in each iteration.

3. The method according to one of claims 1 and 2, wherein Each reference magnetic field carries its information item, and detecting the contribution of the reference magnetic field includes identifying the information of the reference magnetic field.

4. The method according to claim 3, wherein, The information of the reference magnetic field is the characteristic frequency.

5. The method according to claim 4, wherein, The characteristic frequency of the reference magnetic field is the high fundamental frequency of a power of a given number.

6. The method according to one of claims 1 and 2, wherein The magnetometer is a parametric resonance magnetometer, where each parametric resonance magnetometer includes an excitation system (8) that can be enabled to induce a parametric resonance excitation radio frequency field, and in the first stage, the excitation systems of the N magnetometers are enabled.

7. The method according to claim 6, further including a second stage (P2) identical to the first stage (P1), except that the excitation systems of the N - 1 magnetometers other than the measuring magnetometer are disabled.

8. The method according to claim 7, wherein, The second stage includes N iterations, each iteration using a different measuring magnetometer among the N magnetometers.

9. A magnetic field measuring device includes a computer and an array of N vector magnetometers, wherein, Each magnetometer includes a field cancellation system (10) that can be enabled to operate the magnetometer in a zero field. The N magnetometers can perform the following steps while being divided into N - 1 magnetometers with the field cancellation system disabled and one measuring magnetometer with the field cancellation system enabled: o Generating (GENj) by the measuring magnetometer and each of the N - 1 magnetometers with the field cancellation system disabled a plurality of reference magnetic fields with known amplitudes and different directions; o Measuring (MESi) the ambient magnetic field by the measuring magnetometer on a plurality of measurement axes; And the computer is configured to detect the contribution of the reference magnetic field to the ambient magnetic field measurement on one of the measurement axes and determine the coupling coefficient between the measuring magnetometer and one of the magnetometers by calculating the ratio between the amplitude of the contribution of the reference magnetic field and the known amplitude of the reference magnetic field.

Images