Analysis of multi-channel measurement data using orthogonal virtual channels
The present invention introduces a method for processing multichannel measurement data achieved especially in MEG and EEG measurements. The method uses a signal space separation (SSS) method and the orthogonality of lead fields in order to calculate linear transformation from physical measurement channels to virtual channels. The geometry related to the measurement arrangement is dissipated and the number of virtual channels is clearly lower than the number of physical sensors. The concept of total information can be applied for such transformed measurement data due to orthogonality. The method offers simplified post-processing of the biomagnetic data, such as for source modelling. The total information can also be interpreted as a robust quantity describing the physiological state of a patient.
1. A method for spatial processing of multi-channel measurement data implemented on a control electronics, comprising:
measuring multi-channel data with plurality of sensors;
estimating by the control electronics from the measured data, by using a Signal Space Separation method, signals representing biomagnetic sources and signals representing sources outside of a sensor area; and presenting, by the control electronics, the estimated signals as multipole moments having orthogonal lead field expressions, resulting in virtual channel data with dissipated measurement geometry data and with less channels than the number of the plurality of sensors.
2. The method according to claim 1 , wherein in the estimation step is performed with a signal space separation Signal Space Separation method, a magnetic field that was registered using a multi-channel measuring device, is analysed in a geometry in which the sources of interest are disposed in volume V1; the sensors measuring the field or the components thereof in volume V2 outside volume V1; and sources of magnetic interference and compensating actuators in volume V3 outside both volumes V1 and V2, wherein volume V3 can be infinite, wherein the Signal Space Separation method further comprises the steps of:
parametrising in volume V2 the magnetic field produced by the sources of interest as a sum of elementary fields, each of them being irrotational, sourceless and finite outside volume V1 so that a presentation of a desired accuracy is achieved for the parametrised magnetic field in volume V2;
parametrising in volume V2 the sum magnetic field produced by the interference sources and the compensating actuators as a sum of elementary fields, each of them being irrotational, sourceless and finite outside volume V3 so that a presentation of a desired accuracy is achieved for the parametrised magnetic field in volume V2;
calculating the signal vector of the measuring device corresponding to each elementary field;
measuring the magnetic signal using sensors; and
separating the fields produced by sources disposed in different volumes by estimating the components of the measured signal vector in the basis formed by the signal vectors associated with the elementary fields.
3. The method according to claim 2 , the method further comprising:
identifying and processing external interferences with the signal space separation Signal Space Separation method; and
cancelling the external interferences from the measurement data.
4. The method according to claim 2 , the method further comprising:
identifying movement of a patient's head; and
cancelling the effect of the movement of the patient's head with the Signal Space Separation method.
5. The method according to claim 1 , the method further comprising:
estimating the multipole moments with vector spherical harmonic functions.
6. The method according to claim 1 , the method further comprising:
extracting total information from the data represented in virtual channels.
7. The method according to claim 6 , the method further comprising:
calculating the total information I tot with a formula
I
tot
=
1
2
∑
i
=
1
N
log
2
(
SNR
i
+
1
)
.
where SNR i is the power signal to noise ratio of the i:th channel and N is the number of calculated channels.
8. The method according to claim 6 , the method further comprising:
using the total information for estimating the physiological state of the patient, or for identifying and abandoning non-useful virtual channels or for optimizing the Signal Space Separation decomposition results.
9. The method according to claim 1 , the method further comprising:
setting the virtual channels represented by the multipole moments as point-like.
10. The method according to claim 1 , wherein the method is applied in a magnetoencephalographic measurement.
11. The method according to claim 1 , wherein the method is applied in an electroencephalographic measurement.
12. A multi-channel measuring device for spatial processing of multi-channel measurement data, comprising:
a sensor array comprising plurality of sensors for measuring multi-channel data; and
a control electronics configured to estimate from the measured data, by using a Signal Space Separation method, signals representing biomagnetic sources and signals representing sources outside of a sensor area;
the control electronics configured to present the estimated signals as multipole moments having orthogonal lead field expressions, resulting in virtual channel data with dissipated measurement geometry data and with less channels than the number of the plurality of sensors.
13. The device according to claim 12 , wherein the control electronics is configured to perform the Signal Space Separation method, wherein a magnetic field that was registered using the multi-channel measuring device, is analysed in a geometry in which the sources of interest are disposed in volume V1; the sensors measuring the field or the components thereof in volume V2 outside volume V1; and sources of magnetic interference and compensating actuators in volume V3 outside both volumes V1 and V2, wherein volume V3 can be infinite, wherein the control electronics is further configured to perform the steps of:
parametrising in volume V2 the magnetic field produced by the sources of interest as a sum of elementary fields, each of them being irrotational, sourceless and finite outside volume V1 so that a presentation of a desired accuracy is achieved for the parametrised magnetic field in volume V2;
parametrising in volume V2 the sum magnetic field produced by the interference sources and the compensating actuators as a sum of elementary fields, each of them being irrotational, sourceless and finite outside volume V3 so that a presentation of a desired accuracy is achieved for the parametrised magnetic field in volume V2;
calculating the signal vector of the measuring device corresponding to each elementary field;
measuring the magnetic signal using sensors; and
separating the fields produced by sources disposed in different volumes by estimating the components of the measured signal vector in the basis formed by the signal vectors associated with the elementary fields.
14. The device according to claim 13 , the device further comprising:
the control electronics configured to identify and process external interferences with the Signal Space Separation method; and
compensating actuators and the control electronics configured to cancel the external interferences from the measurement data.
15. The device according to claim 13 , the device further comprising:
the control electronics configured to identify movement of the patient's head; and
the control electronics configured to cancel the effect of the movement of the patient's head with the Signal Space Separation method.
16. The device according to claim 12 , the device further comprising:
the control electronics configured to estimate the multipole moments with vector spherical harmonic functions.
17. The device according to claim 13 , the device further comprising:
the control electronics configured to extract total information from the data represented in virtual channels.
18. The device according to claim 17 , the device further comprising:
the control electronics configured to calculate the total information I tot with a formula
I
tot
=
1
2
∑
i
=
1
N
log
2
(
SNR
i
+
1
)
,
where SNR i is the power signal to noise ratio of the i:th channel and N is the number of calculated channels.
19. The device according to claim 17 , the device further comprising:
the control electronics configured to use the total information for estimating the physiological state of the patient, or for identifying and abandoning non-useful virtual channels or for optimizing the Signal Space Separation decomposition results.
20. The device according to claim 12 , the device further comprising:
the control electronics configured to set the virtual channels represented by the multipole moments as point-like.
21. The device according to claim 12 , wherein the device is a MEG measuring device.
22. The device according to claim 12 , wherein the device is an EEG measuring device.
23. A computer program implemented on a non-transitory computer readable medium for spatial processing of multi-channel measurement data by a control electronics, comprising code adapted to perform the following steps when executed on a data-processing system:
measuring multi-channel data with plurality of sensors; and
estimating by the control electronics from the measured data, by using a Signal Space Separation method, signals representing biomagnetic sources and signals representing sources outside of a sensor area; and
presenting, by the control electronics, the estimated signals as multipole moments having orthogonal lead field expressions, resulting in virtual channel data with dissipated measurement geometry data and with less channels than the number of the plurality of sensors.
24. The computer program according to claim 23 , further adapted to perform the estimating step with wherein in the Signal Space Separation method, a magnetic field that was registered using a multi-channel measuring device, is analysed in a geometry in which the sources of interest are disposed in volume V1; the sensors measuring the field or the components thereof in volume V2 outside volume V1; and sources of magnetic interference and compensating actuators in volume V3 outside both volumes V1 and V2, wherein volume V3 can be infinite, wherein the Signal Space Separation_method further comprises the steps of:
parametrising in volume V2 the magnetic field produced by the sources of interest as a sum of elementary fields, each of them being irrotational, sourceless and finite outside volume V1 so that a presentation of a desired accuracy is achieved for the parametrised magnetic field in volume V2;
parametrising in volume V2 the sum magnetic field produced by the interference sources and the compensating actuators as a sum of elementary fields, each of them being irrotational, sourceless and finite outside volume V3 so that a presentation of a desired accuracy is achieved for the parametrised magnetic field in volume V2;
calculating the signal vector of the measuring device corresponding to each elementary field;
measuring the magnetic signal using sensors; and
separating the fields produced by sources disposed in different volumes by estimating the components of the measured signal vector in the basis formed by the signal vectors associated with the elementary fields.
25. The computer program according to claim 24 , further adapted to perform the steps of:
identifying and processing external interferences with the Signal Space Separation method; and
cancelling the external interferences from the measurement data.
26. The computer program according to claim 24 , further adapted to perform the steps of:
identifying movement of the patient's head with the Signal Space Separation method; and
cancelling the effect of the movement of the patient's head.
27. The computer program according to claim 23 , further adapted to perform the step of:
estimating the multipole moments with vector spherical harmonic functions.
28. The computer program according to claim 23 , further adapted to perform the step of:
extracting total information from the data represented in virtual channels.
29. The computer program according to claim 28 , further adapted to perform the step of:
calculating the total information I tot with a formula
I
tot
=
1
2
∑
i
=
1
N
log
2
(
SNR
i
+
1
)
,
where SNR i is the power signal to noise ratio of the i:th channel and N is the number of calculated channels.
30. The computer program according to claim 28 , further adapted to perform the step of:
using the total information for estimating the physiological state of the patient, or for identifying and abandoning non-useful virtual channels or for optimizing the Signal Space Separation_decomposition results.
31. The computer program according to claim 23 , further adapted to perform the step of:
setting the virtual channels represented by the multipole moments as point-like.
32. The computer program according to claim 23 , wherein the computer program is applied in a magnetoencephalographic measurement.
33. The computer program according to claim 23 , wherein the computer program is applied in an electroencephalographic measurement.