WO2012052120A2

WO2012052120A2 - Method and system for determining the position of an apparatus

Info

Publication number: WO2012052120A2
Application number: PCT/EP2011/005051
Authority: WO
Inventors: Volkmar Schultze
Original assignee: Rayonex Schwingungstechnik Gmbh
Priority date: 2010-10-18
Filing date: 2011-10-10
Publication date: 2012-04-26
Also published as: DE102010048574A1; WO2012052120A3

Abstract

The invention relates to a method for determining the spatial position of an apparatus having a rotating magnetic dipole using at least one three-axis magnetic field sensor, wherein the magnetic field emitted by the magnetic dipole is measured by the magnetic field sensor and is evaluated in order to determine the position coordinates in a three-axis coordinate system, wherein at least two individual measurements are carried out and are distinguished by a different relative position between the axis of rotation of the dipole and the position vector which extends between the dipole and the magnetic field sensor, wherein these different relative positions do not result from a movement of the apparatus.

Description

"Method and system for determining the position of a device"

The invention relates to a method for determining the position of a device having a rotating magnetic dipole using at least one three-axis magnetic sensor, wherein the magnetic field emitted by the magnetic dipole is measured by the magnetic field sensor and evaluated to determine the position coordinates in a three-axis coordinate system. The invention further relates to a system for carrying out such a method comprising the device having a rotating magnetic dipole and at least one triaxial magnetic field sensor.

Microsurgical and endoscopic instruments used in medicine are used in particular for diagnostics and operations on sensitive or difficult to access tissues and organs. These interventions are usually computer- and / or camera-controlled and usually require a maximum of precise positioning, positioning and movement of the instruments. For this purpose, probe systems such as magnetic or electromagnetic probes are used. Thus, US Pat. No. 5,836,869 and US Pat. No. 6,248,074 describe magnetic field sources or magnetic field sensors which measure the three spatial coordinates of a moving magnetic field. As a result, however, no spatially accurate or timely position determination of the endoscopic device is possible. This is explained by the fact that the method of determining the magnetic field coordinates described in US Pat. No. 5,836,869 is based on the three-axis, three mutually perpendicular

CONFIRMATION COPY Coils existing magnet makes it necessary to measure three different magnetic fields, which are, to avoid interference, measured in a time-delayed succession by the individual coils are time-delayed applied to electrical voltage. The measurement takes place here outside the patient and also requires a conversion in order to be able to estimate the position of the endoscope in the body.

US 6,248,074 describes the attachment of a magnetic field source outside the patient. The localization takes place here by determining the relative position of the detector to the external magnetic field via a magnetic field sensor attached to the distal end of the endoscope. Again, only a relatively inaccurate measurement is possible because endoscope and sensor are moved relative to the fixed magnetic field and thus no exact relation between the fixed magnetic field coordinates and the changing spatial orientation of the sensor is given. In addition, there are other obstacles and factors that affect the accuracy, such as the problem of measuring in regions far away from the surface of the body, or the impairment of measurement accuracy by external magnetic fields.

An improved method for locating a device as described above is known from WO 2003/103492 A1. Therein it is disclosed to arrange a magnetic dipole in the housing of a medical device or, for example, also of a drill head, which is rotationally driven independently of an optionally occurring rotation of the housing. To locate the device, the magnetic field generated by the magnetic dipole is measured and evaluated by a three-axis magnetic field sensor (fluxgate). This makes it possible to accurately determine the position of, for example, a medical device in the body of a patient. In WO 2003/103492 A1 also discloses a way to determine the roll angle of the housing of the device. For this purpose, a variable component of the magnetic field, which depends on the roll angle, generated, which can be determined by the magnetic field sensor. As a concrete embodiment, a short-term stopping or tilting of the dipole in a defined relative position with respect to the housing is described.

A further development of the method known from WO 2003/103492 A1 is disclosed in DE 10 2005 051 357 A1. There is provided, regardless of the

Device rotatably driven dipole no longer rotate about a rotation axis which is coaxial or parallel to the axis of rotation of the self-rotating device driven, but obliquely thereto. This is intended to ensure that the position of the magnetic field generated by the magnet is set in an evaluable relative dependence on the housing. This can, despite the

Independence of the rotation of the device from that of the arranged therein magnetic dipole, be closed by an evaluation of the magnetic field on the roll angle of the device.

Both in the method according to WO 2003/103492 A1 and in DE 10 2005 051 357 A1 a problem remains unresolved. The localization of the magnet via measurement and evaluation of the magnetic field emitted by it corresponds to the solution of a so-called "inverse problem." As an inverse problem, the problem is called to infer from the effects of a system on the cause In general, an inverse problem is not clearly solvable.

In the method known from WO 2003/103492 A1, the inverse problem is to reconstruct the location and the orientation of the magnetic dipole from the magnetic field of the rotating magnetic dipole. This is not readily possible, because in a merely mathematical evaluation of the measured magnetic field results for the retroactive calculation of the location of the magnet four equivalent solutions. Without additional information, it is therefore not possible to select the right one from the four solutions offered. Therefore, for carrying out the method known from WO 2003/103492 A1, it is necessary, for example by means of a visual estimation, to define boundary conditions by means of which the four mathematical solutions can be reduced to one. However, this is not possible in particular in the applications mentioned in WO 2003/103492 A1, namely the localization of an endoscopic medical instrument or a horizontal drilling device, since there is no visual contact with the instrument or the drilling device. A further problem is that the selection of the correct solution depends on how the axis of rotation of the magnet and the position vector extending between the magnetic dipole and the magnetic field sensor are oriented relative to one another. Thus, this selection changes when the angle between these two factors (axes of rotation and position vector) passes through 90 ° (ie from <90 ° to> 90 ° alternating or vice versa). The approach of initially selecting a correct solution as boundary condition by sight, and then extrapolating from the course of the localized locations of the magnetic dipole into the future and seeing which of the offered solutions comes closest to this extrapolation, can in many situations lead to wrong results.

The invention was therefore based on the object of specifying an improved method and a corresponding system for determining the spatial position of a device. In particular, a simpler and / or more accurate localization should be achieved. This object is achieved by a method according to independent claim 1 and by a system according to independent claim 7. Advantageous embodiments are the subject of the respective dependent claims and will become apparent from the following description of the invention.

The core of the method according to the invention is to perform at least two individual measurements for a location measurement, which are characterized by a different relative position between the axis of rotation of the dipole and the position vector extending between the dipole and the magnetic field sensor, these different relative positions not from a possible movement of the device result, ie are independent of it.

The choice of the right solution then results from the fact that the two separate measurements make it impossible for the right solution - or the wrong one - to do so to a lesser extent (in case of movement of the device)

Magnetic field sensor and / or the device itself) changed.

Since in the (real) application of the method according to the invention, the measured values of the magnetic field sensor are superimposed by noise, in the defined relative change in position usually not twice (or several times) exactly the same

Location to be determined. The more general selection criterion for determining the correct solution is therefore to select the solution whose location changes less with the change in position. The solution selection according to the invention, i. to determine which solution has its value

(i.e., the determined position) in a comparison of different relative positions of rotation axis and position vector does not leave or only relatively little leaves no extrapolation of future solutions based on already determined solutions. A corresponding system according to the invention comprises, in addition to a device comprising a rotating magnetic dipole and (at least) a triaxial magnetic field sensor, i. a magnetic field sensor, which can determine the strength (magnetic flux density) of a magnetic field in at least three different coordinates, in addition to means for the performance of at least two individual measurements, wherein the individual measurements by a different relative position between the

Rotation axis of the dipole and the extending between the dipole and the magnetic field sensor position vector, wherein the different relative positions do not result from a movement of the device.

In a preferred embodiment of the method according to the invention, the different relative positions between the axis of rotation and the position vector achieved in that the position of the axis of rotation of the magnetic dipole is changed for the two individual measurements. This embodiment has the particular advantage that as a result the positions of the magnetic dipole (apart from a possible movement of the device itself) and the sensor do not change globally, and thus the selection of the correct solution according to the invention is easily verifiable.

Structurally, this can be implemented by the dipole in a system according to the invention, for example, is rotatable about an additional second axis of rotation, which is arranged obliquely with respect to the first axis of rotation of the dipole. This superposition of the two rotations of the magnetic dipole results in a cyclic (with a constant rotation about both axes of rotation) change in position of the (first) axis of rotation to the position vector, which according to the invention can be used to select the correct solution. For the solution selection already an inclination of the two axes of rotation of the magnetic dipole ranges by a few degrees (preferably at least 3 °), the accuracy with which a selection can be made, with increasing inclination increases. The inclination can reach up to an angle of 90 °, in which the (first) axis of rotation of the magnetic dipole rotates in a circle.

A rotation of the dipole about an additional second axis of rotation oriented obliquely with respect to the first axis of rotation can be achieved on the device side by rotating the device within which the magnetic dipole is arranged itself about a rotation axis and thus representing this axis of rotation as the second axis of rotation.

Another possibility for achieving a defined change in the relative position between the axis of rotation of the dipole and the position vector can be achieved in that the position of the magnetic field sensor (possibly additionally by a change in position due to the movement of the device) between the two individual measurements relative to the Device is changed, and preferably along a defined path or between two or more defined positions. This can be done on the device side by suitable means for changing the position of the magnetic field sensor relative to the device between at least two defined positions. This solution, in which the relative position change by a defined

Changing the location of the magnetic field sensor is achieved, the selection of the correct solution in turn allows the fact that the correct solution than that, the result of which despite the change in position does not or only to a lesser extent changes. Of course, the defined change in position of the magnetic field sensor must be taken into account and mathematically corrected. The magnetic field sensor may then be continuous between the at least two defined positions are moved (eg by an alternating or rotating movement), whereby a regular execution of the solution selection can be done. Alternatively, the movement can only take place as needed. A relative defined change in position of the one magnetic field sensor can be achieved, for example, by displacing it in a holding device between two defined end points. This can be done both translationally and rotationally (for example by means of a hinge).

According to the invention, a "defined relative change in position" refers to the respective relative positions of the axis of rotation of the dipole and the position vector extending between the dipole and the magnetic field sensor during the two individual measurements of a position measurement Single measurements in the above embodiments occur temporally successively and the defined relative change in position is generated by an actual change in position of the dipole or the magnetic field sensor, it is also possible to achieve the different relative positions in that the magnetic field (also simultaneously) measured by at least two magnetic field sensors The selection of the correct solution then results from the fact that the positions, which are respectively determined by means of the two magnetic field sensors, at the wrong solution ng have a greater distance from each other than the right ones. The accuracy of the solution selection increases with the distance between the two magnetic field sensors, whereby even at a distance of 10 cm (and below) usable results can be achieved.

In an advantageous embodiment of an embodiment of the invention based on the use of two magnetic field sensors, it is provided to arrange the two magnetic field sensors in the direction of gravity one above the other. This can in particular have handling and design advantages. Under

Gravity direction is to be understood here in the direction of gravitational acceleration or in the direction of gravity of the earth.

Since it may be difficult to align the magnetic field sensors exactly in the gravitational direction when using the system according to the invention, this may preferably further comprise an inclination sensor with which the inclination of the connecting line extending between the magnetic field sensors relative to the gravitational direction can be determined. This makes it possible to compensate for the (incorrect) localization errors caused by the non-exact alignment in the direction of gravity. In a particularly preferred embodiment of the method according to the invention, provision may be made for the magnetic field sensor to be positioned relative to the device such that the sign of the coordinate of a selected axis of the coordinate system is uniquely determined. This should preferably be done before or at the beginning of the measurement. As a result, two of the four solutions, which result mathematically from an evaluation of the magnetic field emitted by the magnetic dipole, can be eliminated. The remaining two solutions can then be evaluated according to the invention by the defined change in the relative position of the axis of rotation of the dipole and the position vector and thus the correct solution can be determined.

If, due to a movement of the device, the coordinate of the selected axis passes through zero, there is a sign change not only of this, but also of the other coordinates; in these even jumpy. This erratic change can be used to computationally change the sign in all

Coordinates to perform the previous positioning measurement. This can be changed without delay to the then correct, in the previous measurements still wrong of the two solutions. Preferably, the inventive method or the inventive

System for localization of the spatial position of a medical and in particular microsurgical or endoscopic instrument used. However, it is not limited to this application, but is in principle suitable for the localization of any, in particular difficult to access and / or not visible device. For example, here Erdarbeitsvorrichtungen and horizontal drilling devices are called.

Under a "Erdarbeitsvorrichtung" according to the invention any device for creating holes, for expanding holes and for pulling pipes or pipes in holes or old pipes understood within the soil.

According to the invention, "soil" is understood to mean any accumulation of a material or material mixture into which a bore can be introduced, including in particular not only the soil per se, but also any bulk material deposits on the earth's surface, such as building material fill.

The invention will be explained in more detail with reference to embodiments shown in the drawings.

In the drawings shows: Fiq. 1 shows a system according to the invention in a first embodiment;

Fiq. 2 shows a system according to the invention in a second embodiment;

Fiq. 3 shows a system according to the invention in a third embodiment;

Fiq. 4 the physical quantities relevant for determining the position of a device of a system according to the invention and their relationships; and

5a to 9 are graphical representations of the results of a series of measurements for

Clarification of the method according to the invention.

Fig. 1 shows a schematic representation of an inventive system for locating a device.

The device is, for example, a medical instrument for endoscopic or microsurgical applications or a drill head of a horizontal drilling device.

Within the device, a magnetic dipole 1 (in the present case a permanent magnet) is rotatably mounted. The magnetic dipole 1 is connected via a shaft 2 to an electric motor 3, which drives it in rotation. Alternatively, for example, a hydraulic or pneumatic drive, such as corresponding turbines, may be provided. The axis of rotation 4 of the dipole 1 is oriented obliquely with respect to the longitudinal axis 5 of the device. The rotating dipole 1 generates a magnetic field which rotates with it. The magnetic field can be measured by a receiving unit 6, for example arranged on the earth's surface, which comprises at least one three-axis magnetic field sensor 7 and evaluated to determine the position of the magnetic dipole 1. Starting from a fixed coordinate system of the receiving unit 6 and the magnetic field sensor 7, the rotating magnetic field of the dipole 1 is as a changing with respect to the size and direction magnetic field vector. Specifically, can be determined by means of the magnetic field sensor 7, the rotating magnetic field vector whose origin is the position of the rotating dipole 1 defined. By evaluating the time-varying magnetic field emitted by the dipole 1 by the receiving unit 6, the position of the drill head located in the ground can thus be determined, whereby this determination first mathematically produces four different solutions. Of these solutions, each have two identical absolute coordinates in all coordinates, but their signs are opposite. This results in two pairs of solutions, which are symmetrical to each other to the origin of the coordinate system. From these four solutions, a clear solution is determined by the method according to the invention. For this purpose, the "correct" of the two pairs of solutions is first selected by positioning the device and the magnetic field sensor at the beginning of the measurement or localization in such a way that the measured value is given a sign that is as accurately defined as possible relative to a coordinate of the three-axis coordinate system This allows the coordinate space to be divided into two "hemispheres" that differ (clearly) in the sign of one of the three coordinates. From the knowledge of the approximate position of the device, the "correct" sign can be determined, allowing to select and exclude the "wrong" of the two pairs of solutions. Then for the unambiguous determination of the (only) correct solution, only one determination of the "right" solution from the remaining solution pair is required.

This is done according to the invention by at least two individual measurements, which are characterized by different relative positions between the axis of rotation of the dipole and the position vector 10, which extends between the origin of the coordinate system of the magnetic field sensor 7 and the dipole 1, wherein the different relative positions not one (possible ) Movement of the device itself.

When computationally determining the correct solution of the remaining pair of solutions can lead to a sign change, if by a movement of the

Device, the coordinate used for the "hemispherical selection" (in this case x) passes through zero, ie if the position vector and the rotation axis are perpendicular to each other, but since the two corresponding ones of the four solutions are always mirror-symmetrical to the coordinate origin, then all coordinates of these two always have Solutions of opposite sign Therefore, the interaction of a change of the hemisphere (which is equivalent to the sign change of x) and a clipping of the sign of x means that the signs of the other coordinates (here y and z) are immediately wrong Therefore, this would not be the case theoretically only if all the coordinates were changed when changing the

Hemisphere would go through zero, but this is not or only with difficulty, because the magnet would then have to cross the sensor. This jump can be determined unambiguously if, for example, the criterion K =

-, y _M | ⁺ | ^z _< - z _f _, | used, where i is the current

Number of the respective location measurement is. Thus, the difference between the y and z coordinates of two successive positioning measurements (each with two

Single measurements). In the hemispheric change, K has a significantly greater value than in the entire remaining course of the location measurements. The occurrence of such a "peak" can thus be used as a request to change the signs in all coordinates with respect to the previous positioning measurements, whereby the criterion K is independent of the direction, and thus also becomes a possible reversal of the direction of the device to be located and a possibly renewed passing of the "hemispherical boundary" again force a change of the sign again 1 embodiment achieved by a superposition of two rotational movements of the magnetic dipole 1, namely on the one hand the rotation of the dipole 1 about its own axis of rotation 4 and on the other by a (co-) rotation of the dipole 1 arranged inside the device about the axis of rotation The position of the rotation axis 4 is thereby changed continuously and cyclically relative to the position vector 10. The embodiment according to FIGS. 2 and 3 differ from that of FIG

1 in that not the position of the axis of rotation 4 of the rotating dipole 1 is changed, but the position vector 10 itself is displaced in a defined manner between the two individual measurements by either changing the (global) position of the magnetic field sensor 7 in a defined manner (FIG. 2). or two magnetic field sensors 7a, 7b are used (FIG. 3), resulting in two position vectors 10a, 10b which have different relative positions to the axis of rotation 4 of the dipole 1. In both cases, a recalculation of the new or second position to the old or first, in order to compensate for the influence of the change in position of a magnetic field sensor 7 and the distance between the two magnetic field sensors 7a, 7b on the solution selection.

The change in position of the one magnetic dipole is achieved in the embodiment according to FIG. 2 in that the magnetic field sensor 7 can be moved in a frame 8 of the receiving unit 6 along a guide between two defined positions (see double arrow). The method of the magnetic field sensor 7 is carried out according to demand either cyclically continuously or at defined times.

The two magnetic field sensors 7a, 7b in the embodiment according to FIG. 3 are arranged one above the other within a frame 8 of the receiving unit 6 (in the direction of gravity). Since this alignment (one above the other) does not always have to be exact, but at the same time has an influence on the accuracy of the solution selection, the receiving unit additionally has an inclination sensor 9, which detects a possible deviation from the vertical alignment, so that a corresponding computational compensation takes place can. FIG. 4 shows a schematic diagram of the physical quantities required for determining the position of the magnetic dipole and their relationships.

Where m represents the magnetic dipole moment at the location of the dipole (origin of the coordinate system), B is the flux density generated by this at the location of the sensor. The vector D (position vector) points from the dipole to the sensor. Θ is the angle between m and D. x is the angle between D and B. n is the normal vector of the ellipse plane and v is the vector pointing in the direction of movement of the device (and thus of the dipole) (propulsion vector).

FIGS. 5a to 9 show the results of a measurement series which further clarify the mode of operation of the solution selection according to the invention. The measurement series is based on the principle of using two magnetic field sensors (see Fig. 3). The measurements were carried out along a straight measurement path which runs approximately parallel to the x-axis of the coordinate system of a stationary magnetic field sensor. A device with a rotating magnetic dipole was moved along the measuring path, the steps becoming smaller and smaller with increasing proximity to the magnetic field sensor. The x-coordinate was varied between ± 11 m and the distance in the y-direction was about 70 cm. The magnetic field sensor was sequentially mounted at four different heights for each position of the magnetic dipole to determine which distance between two magnetic field sensors is favorable for good discrimination of the correct solution. The lowest height was about 60 cm above the measuring section (height z); the other three heights of the magnetic field sensor were each 30 cm above the previous one. In the evaluation, these positions are referred to as P1 (height z), P2 (height z + 30cm), P3 (height z + 60cm) and P4 (height z + 90cm).

FIGS. 5a to 5d show the composition of the measurement results with these four positions of the magnetic field sensor. It can be seen that the two solutions with the correct sign of the x-coordinate (in this case solutions 3 and 4) exchange the roles when passing through the magnetic field sensor (ie at x = 0). For x <0 solution 4 is the right one, for x> 0 it is solution 3. The wrong solution moves extremely quickly away from the correct path of the measurement path.

In FIGS. 6a and 6b, the measurement results with the solutions 3 and 4 for the four magnetic field sensor heights are shown together after the above-mentioned height offset has been eliminated. It can be seen that the locations coincide with the correct solution for all four magnetic field sensor heights, but differ significantly in the wrong solution. This is again summarized in FIGS. 7a to 7c as one would interpret this practically in the case of two magnetic field sensors arranged one above the other. It calculates the geometric distance of the determined with two superimposed magnetic field sensors locations, namely for the two available solutions. It is clear that for the right solution this distance is zero, while it stands out clearly for the wrong solution. Of course, this is the clearer the greater the distance between the two magnetic field sensors.

The measurement results clearly show that the criterion of the lower geometric distance of the location with two superimposed sensors represents a very clear and infallible criterion, especially for the solution selection when passing the sensor, ie when the two solutions exchange their roles as correct and incorrect. However, it can also be seen that for long distances, where locating accuracy decreases, this criterion can be affected by poor locating results (Figure 7c). Here, the safety of the selection becomes better, the larger the geometric distance of the sensors is (P4-P1) gives clearer results than (P3-P1) or even (P2-P1)).

In the location method according to the invention, each locating result stands alone. Should the wrong solution be chosen due to any error, this has no effect on the next location. This is easy to see, considering the worst choice for solution selection. Of course, this is the measurement with the smallest distance between the sensors, ie (P2-P1) in FIG. 7c. There, at the greatest distances on both sides, there is the case that the solution distance of the wrong solution is lower than that of the right one.

Fig. 8 shows the location and the solution selection, as resulting from the criterion of the lower geometric distance. You can see the "slip-up" of the wrong solution, but you can also see that the further tracking of it is unaffected. In Fig. 7d, in addition to the location of the two magnetic field sensors and their mean value is given. This increases the accuracy of the location a bit more, since twice as many measurements are received. The two magnetic field sensors can also be used to increase the quality of the measurements.

The location according to the invention by means of two magnetic field sensors can provide further information: So far, the geometric distance of the solutions was considered only for the selection of the correct solution. But you can also use the size of the distance of the right solutions as a criterion for the quality of the measurement, because with really good measurements, it must be close to zero. Fig. 9 shows a corresponding result. It will now only the detection results with sufficient Goodness offered. So you have a reliable selection option, where you can set the quality criterion on the allowable distance of the solutions of the two magnetic field sensors used. FIG. 9 also shows that in the method according to the invention described here

Area around x = 0, where neither of the two solutions is really correct, can be extremely narrowed. The erroneous deviation there is in the range of 5 cm.

Claims

claims:

Method for determining the spatial position of a device comprising a rotating magnetic dipole (1) using at least one three-axis magnetic field sensor (7), the magnetic field emitted by the dipole (1) being measured by the magnetic field sensor (7) and determining the position coordinates in a three-axis coordinate system is evaluated, characterized in that at least two individual measurements are performed, which is characterized by a different relative position between the axis of rotation (4) of the dipole (1) and between the dipole (1) and the magnetic field sensor (7) formed position vector ( 10), these different relative positions not resulting from movement of the device.

A method according to claim 1, characterized in that between the two individual measurements, the position of the axis of rotation (4) is changed.

A method according to claim 2, characterized in that the dipole (1) is additionally rotated about a second, with respect to the first axis of rotation (4) oriented rotational axis (5).

A method according to claim 1, characterized in that between the two individual measurements, the position of the magnetic field sensor (7) is changed relative to the device.

A method according to claim 1, characterized in that the two individual measurements by means of two magnetic field sensors (7a, 7b) are performed.

Method according to one of the preceding claims, characterized in that the magnetic field sensor (s) (7) is / are positioned relative to the device such that the sign of the coordinate of a defined axis of the coordinate system is uniquely determined.

A method according to claim 7, characterized in that, if by a movement of the device, the coordinate of the defined axis is zero, a mathematical change of the sign of all coordinates is performed in comparison to the previous measurement.

8. System for determining the spatial position of a device comprising the device having a rotating magnetic dipole (1) and at least at least one three-axis magnetic field sensor (7), characterized by means for performing at least two individual measurements, which are characterized by a different relative position between the axis of rotation (4) of the dipole (1) and the (10) between the dipole (1) and the magnetic field sensor (7) distinguished local vector (10), wherein these different relative positions do not result from a movement of the device.

9. System according to claim 8, characterized in that the dipole (1) about a second, obliquely with respect to the first axis of rotation (4) oriented rotation axis (5) is rotatable.

10. System according to claim 9, characterized in that at least the dipole (1) comprising part of the device about the second axis of rotation (5) is rotatable.

11. System according to claim 8, characterized by means for changing the position of the magnetic field sensor (7) relative to the device between at least two defined positions.

12. System according to claim 8, characterized by two triaxial magnetic field sensors (7a, 7b).

13. System according to claim 12, characterized in that the two magnetic field sensors (7a, 7b) are arranged at a distance of at least 10 cm.

14. System according to claim 13, characterized in that the two magnetic field sensors (7a, 7b) are arranged one above the other in the direction of gravity.

15. System according to claim 14, characterized by an inclination sensor (9) for determining the relative orientation of the two magnetic field sensors (7a, 7b) with respect to the gravitational direction.