Method and device for using a multi-channel measurement signal in source modelling
A method for interpreting the current distribution of an object being measured using basis vector components calculated from the measured signals. The components in question have been so selected that they describe the features, as independent as possible, of the current distribution being examined, which enhances the computation and makes it more accurate. This is achieved by converting the measured signals into a more natural form from the standpoint of the current distribution while totally eliminating the signals associated with the external interferences. After the conversion, the source modeling is performed in an optimal manner using the basis vector components of the signal space instead of the actual measurement signals. One substantial feature of the invention is that after the conversion, the source model need not be regularized any more.
1. A method for determining a current distribution of an object, the method comprising:
measuring the magnetic fields in vicinity of the object using a multi-channel measurement device that measures an irrotational and sourceless vector field, whereby one measurement sensor corresponds to each channel;
converting a multi-channel measurement signal corresponding to each measurement sensor into signals of a predetermined set of virtual sensors, which signals are mutually orthogonal; and
determining the current distribution of the object being measured from the signals of the set of virtual sensors in a predetermined function basis.
2. The method according to claim 1 , wherein the object is approximated using a conductor, and a multi-pole expansion of the field is calculated from the multi-channel measurement signal.
3. The method according to claim 2 , wherein the multi-pole expansion is calculated by taking into account magnetic fields emitted by sources outside the object.
4. The method according to claim 2 , wherein the multi-pole expansion is calculated by ignoring magnetic fields emitted by sources outside the object.
5. The method according to claim 4 , wherein:
an orthonormal function basis is placed into a current distribution equation; and coefficients of the current distribution are analytically solved from the equation:
C
lm
=
γ
^
l
M
lm
[
∫
0
R
r
l
+
2
f
l
(
r
)
ⅆ
r
]
-
1
,
wherein c lm are said coefficients, {acute over (γ)} l is a constant associated with order l, M lm are multi-pole coefficients, R is a radius of a sphere to be examined, r is the radial distance as a variable and f l (r) is a selectable radial function.
6. The method according to claim 2 , wherein external interferences are eliminated prior to the step of converting.
7. The method according to claim 2 , wherein as an orthonormal function basis, a current distribution equation of the following form is selected:
J
->
(
r
->
)
=
∑
l
=
0
L
∑
m
=
-
l
l
c
lm
f
l
(
r
)
X
->
l
m
(
θ
,
φ
)
,
wherein J (7) is the current distribution, L and l are orders, C lm are coefficients of the current distribution, f l (r) is a selectable radial function and X lm (θ,φ) is vector spherical harmonic.
8. The method according to claim 7 , wherein function f l (r) is used to adjust a depth weighing of a current distribution model.
9. The method according to claim 1 , wherein the object is approximated using a spherically symmetric conductor.
10. A measurement device for determining a current distribution of an object by measuring magnetic fields in a vicinity of the object, the measurement device comprising:
a set of measurement channels ( 1 , 1 1 , 1 2 , . . . 1 n ) that measure an irrotational and sourceless vector field, in which case at least one measurement sensor 2 , 2 1 , 2 n , . . . 2 n ) corresponds to each channel;
processing means for processing a measurement signal in which the object is approximated using a spherical-symmetrical conductor, wherein
the processing means include a conversion module for converting a multi-channel measurement signal corresponding to each measurement sensor into signals of a predetermined set of virtual sensors, which signals are mutually orthogonal; and
calculation means for determining the current distribution of an object being examined from the set of virtual sensors using depth r in a predetermined orthonormal function basis.
11. The measurement device according to claim 10 , wherein the calculation means are arranged to calculate a multi-pole expansion from the multi-channel measurement signal.
12. The measurement device according to claim 11 , wherein the multi-pole expansion is calculated by taking into account magnetic fields emitted by sources outside the object being measured.
13. The measurement device according to claim 11 , wherein the multi-pole expansion is calculated by ignoring magnetic fields emitted by sources outside the object being measured.
14. The measurement device according to claim 13 , wherein
the orthonormal function basis is placed into the current distribution equation; and
coefficients of the current distribution are solved analytically from the equation:
c
l
m
=
γ
^
l
M
lm
[
∫
0
R
r
l
+
2
f
1
(
r
)
ⅆ
r
]
-
1
,
wherein c lm are said coefficients, {acute over (γ)} l is a constant associated with order l, M lm are multi-pole coefficients, R is a radius of a sphere to be examined, r is the radial distance as a variable and f l is a radial function to be selected.
15. The measurement device according to claim 11 , wherein as the orthonormal function basis, a current distribution equation with the following form is selected:
J
->
(
r
->
)
=
∑
l
=
0
L
∑
m
=
-
l
l
c
lm
f
l
(
r
)
X
->
l
m
(
θ
,
φ
)
,
wherein J ({right arrow over (r)}) is the current distribution, L and l, are orders, c lm are coefficients of the current distribution, f l (r) is a radial function to be selected and X lm (θ,φ) is vector spherical harmonic.
16. The measurement device according to claim 15 , wherein f l (r) is used to adjust a depth weighing of a current distribution model.
17. The measurement device according to claim 10 , wherein the measurement device converts the signals into a set of virtual sensors prior to storage, and analysis software converts the stored data into a current distribution.