Quantitative magnetoencephalogram system and method

• US 6,195,576 B1
• Filed: 03/09/1998
• Issued: 02/27/2001
• Est. Priority Date: 03/09/1998
• Status: Expired due to Term

• Current Assignee: New York University
• Original Assignee: New York University
• Inventors: John, Erwin Roy
• Primary Examiner(s): Lateef, Marvin M.
• Assistant Examiner(s): MANTIS MERCADER, ELENI M
• Application Number:
  US09/036,933
• Time in Patent Office: 1,086 Days
• Field of Search: 600/409, 600/407, 324/244, 324/248, 324/260, 324/261, 324/246
• US Class Current: 600/409
• CPC Class Codes: A61B 5/245 specially adapted for magnetoencephalographic [MEG] signals

Abstract

A magnetoencephalogram (MEG) method and system for detecting and analyzing the magnetic effects of brain waves uses a dewar having a helmet-shaped cavity and an array of magnetic field detectors arranged in an array. In one embodiment the field detectors are located in the same relative locations as electrodes in the EEG 10/20 system. A subject'"'"'s MEG data is compared to a normative spontaneous and/or evoked MEG database collected using standardized detector positions and sensory stimuli and/or to a normative QEEG database simultaneously or previously constructed using corresponding standardized electrode positions and sensory stimuli. Statistical parameters in these databases assess the probability that MEG features derived from a subject are within normal limits. Deviations are indicated on interpolated statistical probability maps color-coded to indicate degree of abnormality. Multivariate statistical analyses may be used to categorize brain disorders in individual patients.

Claims

1. The method of detecting and analyzing magnetic fields on the scalp of the subject arising from the subject'"'"'s brain electromagnetic activity, by:

(a) detecting and amplifying the magnetic fields using a plurality of magnetic field detectors positioned adjacent selected areas of the subject'"'"'s scalp to produce amplified signals in a plurality of channels;

(b) converting the amplified signals into digital MEG (Magnetoencephalogram) data;

(c) automatically comparing, in a computer, the digital MEG (Magnetoencephalogram) data with a normative QEEG (Quantitative Electroencephalogram) reference database of features simultaneously or previously collected from normal subjects, the comparison being on the basis of the same features obtained at the same location of the field detectors, or a computed multiple regression equation, canonical correlation, or covariance of multiple MEG detectors for the subject, as the EEG electrodes positioned on the normal subjects which generated the QEEG database, to obtain a set of QMEG (Quantitative Magnetoencephalogram) variables.

2. A method as in claim 1 wherein the detecting of the analog magnetic fields is accomplished by positioning a helmet-shaped dewar tail over the scalp of the subject, aligning the subject'"'"'s head with fiducial marks on the dewar tail, and holding the subject'"'"'s head firmly in the dewar tail to prevent head movement.

3. A method as in claim 1 and selecting the areas on the subject'"'"'s scalp corresponding in location to areas which at least some of EEG electrodes of the 10/20 EEG system are normally connected.

4. A method as in claim 3 wherein the digital data is derived from 19 magnetic field detectors and is compared to data derived from 19 EEG electrodes corresponding in location to the full 10/20 system.

11. The methods of claim 1, 5 or 10 in which the normative reference databases of features are in the form of age-corrected scores which are in the appropriate dimensional units and all features from each subject and in the databases have been transformed to achieve Gaussianity.

13. A method as in claims 1, 5 or 10 and employing a SQUID as part of each magnetic field detector.

14. A method as in claims 1, 5 or 10 and employing a magnetometer or gradiometer as part of each magnetic field detector.

15. A method as in claims 1, 5 or 10 wherein the digital data derived from each magnetic field detector is log or otherwise transformed for Gaussianity and/or age-regressed and/or Z transformed or F ratio transformed or otherwise statistically assessed, based upon the normative reference database.

5. A method of detecting and analyzing magnetic fields on the scalp of a subject arising from the subject'"'"'s brain waves, by:

(a) positioning or selecting a plurality of magnetic field detectors arranged in an array in a dewar tail;

(b) bringing the dewar tail into close proximity to the subject'"'"'s scalp and holding the subject'"'"'s head at the dewar tail to prevent head movement;

(c) using the magnetic field detectors to detect scalp magnetic fields and to generate analog signals therefrom;

(d) amplifying and digitizing the analog signals to produce a set of digital MEG (Magnetoencephalogram) data corresponding to the analog signals from each magnetic field detector;

(e) automatically, in a computer, statistically comparing features extracted from each set of digital MEG data taken under selected conditions or stimuli with mean values and SD'"'"'s (Standard Deviations) in a normative MEG database, on a feature-by-feature basis, based upon QMEG (Quantitative Magnetoencephalogram) data from a normal group simultaneously or previously gathered under the same conditions or stimuli and with the brain activity MEG detectors of the normal group at the same locations as those of the subject, to obtain and assess QMEG values;

(f) simultaneously with (b), using a plurality of EEG electrodes on the scalp of the subject to derive EEG signals;

(g) amplifying and digitizing the EEG signals to produce sets of QEEG (Quantitative Electroencephalogram) data; and

(h) automatically in the computer comparing each set of QEEG data with digital QEEG data from a normal group taken, simultaneously or previously, under the same conditions or stimuli.

6. A method as in claim 5 and also comparing quantitative features extracted from each set of digital MEG data with mean values and standard deviations of features extracted from digital data of a normal group which is derived from EEG signals.

7. A method as in claim 5 and positioning a plurality of the EEG electrodes both in the 10/20 position and aligned with the magnetic field detectors.

8. A method as in claim 5 and using thin and magnetically transparent EEG electrodes for the plurality of EEG electrodes.

9. A method as in claim 5 wherein the subject'"'"'s digital data is derived from 19 magnetic field detectors and quantitative features derived therefrom are compared to similar features derived from data from 19 EEG electrodes corresponding in location to the full 10/20 system.

10. A method of detecting and analyzing magnetic fields on the scalp of a subject generated by the subject'"'"'s spontaneous or stimulus evoked brain activity, by:

(a) positioning an array of a plurality of magnetic field detectors closely adjacent to the subject'"'"'s scalp to detect and amplify the magnetic fields from selected areas of the scalp;

(b) converting the detected magnetic field into sets of digital data, with each set being generated by a magnetic field detector;

(c) automatically comparing, in a computer, the sets of digital data with a normative reference database of QEEG (Quantitative Electroencephalogram) features simultaneously or previously collected from normal subjects, the comparisons being on a feature-by-feature basis, and the normal subjects having generated the database using brain wave EEG detectors at the same relative locations on their scalps as the relative locations of the magnetic field detectors on the scalp of the subject, to compute a measure of the probability that the observed value is beyond normal limits (Z- or F-score);

(d) deriving a discriminant score based on the comparisons of (c) by selecting, weighting and combining a number of said feature-by-feature comparisons;

(e) combining the discriminant scores to derive probabilities that the subject is classed in a specific diagnostic category;

(f) applying selected guardbands (rule-out levels) to the probabilities to classify the subject on the basis of the probabilities; and

(g) displaying the classification of the subject.

12. A method as in claim 10 and selecting the areas on the subject'"'"'s scalp which correspond in location to areas to which at least some of the electrodes of the 10/20 EEG system are normally connected.

16. A method as in claim 10 wherein the subject'"'"'s digital data is derived from 19 magnetic field detectors and quantitative features derived therefrom are compared to similar features from data derived from 19 EEG electrodes corresponding in location to the full 10/20 system.

17. A method as in claim 10 wherein the detecting of the analog magnetic fields is accomplished by positioning a helmet-shaped dewar tail over the scalp of the subject, aligning the subject'"'"'s head with fiducial marks on the dewar tail, and holding the subject'"'"'s head firmly in the dewar tail to prevent head movement.

18. The method of analyzing the brain waves of a human subject to determine the presence, absence, degree and type of magnetic evoked response abnormality, the method consisting of the steps, in sequence, of:

(a) positioning a plurality of magnetic field detectors in an array proximate to the scalp of the subject, the detectors being positioned in the same relative positions over the scalp as at least some of the electrodes are positioned in the 10/20 system;

automatically presenting a sequence of stimuli to the subject to evoke magnetic brain responses (MER) at the detectors;

(b) detecting, amplifying, converting the amplified electromagnetic brain activity from each detector into digital data, averaging the digitized brain activity to provide a set of averaged magnetic evoked responses (AMERs) from the array;

(c) decomposing each of the averaged magnetic evoked responses (AMER) by factor analysis, in which each AMER from each detector is compared to pre-formulated factor waveforms (or component shapes) stored in computer memory to determine the amount of power that each pre-formulated factor waveform contributes to said AMER to thereby produce a factor score a_ij, quantifying the contribution of each factor j to the AMER from any detector i;

(d) subjecting the results of the said factor score to Z transform, wherein each of said factor scores is compared to the statistical parameters of a normative factor score, corresponding to a quantitative assessment of AMER morphology, thereby providing an estimate of the degree of abnormality of each of said AMER'"'"'S.

19. The method of claim 18 and displaying the said normality and degree of abnormality on a color-coded sectored or interpolated topographic map comprising a head diagram in which each area of the head represents a detector location and the colors represent normality and the degree of abnormality reflected by each factor Z score.

20. The method of claim 19 wherein a series of topographic maps correspond to the anatomical distribution of Z-scores for each factor, a map indicating the % of local AMER variance which cannot thereby be accounted for (residual), and a map depicting the square root of the sum of Z-scores at each detector $|\sum _{}^{}$

i=iM

 



Zi2

|1/2

or total abnormal morphology.

21. The method of claim 18 and positioning 19 detectors in the same relative positions as the 19 active electrodes in the 10/20 EEG system.

22. The method of claim 18 wherein the stimuli presented are a series of flashes or other visual stimuli to produce visual magnetic evoked responses (VMERs).

23. The method of claim 18 wherein the stimuli presented are a series of clicks or other auditory stimuli to produce auditory magnetic evoked responses (AMERs).

24. The method of claim 18 wherein the stimuli presented are a series of electrical shocks or other somatosensory stimuli to produce somatosensory magnetic evoked responses (SMER'"'"'s).

25. A system for detecting and analyzing the magnetic fields on the scalp of the subject arising from the subject'"'"'s brain activity, comprising:

(a) means for detecting and amplifying analog magnetic fields including a plurality of magnetic field detectors adapted to be positioned adjacent selected areas of the subject'"'"'s scalp to produce amplified signals in a plurality of channels;

(b) means for converting the amplified signals into digital data;

(c) computer means for automatically comparing quantitative features extracted from the digital data with statistical parameters of a feature database based on EEG (Quantitative Electroencephalogram) of normal subjects, the comparison being on the basis of the same features obtained at the same location of field detectors for the subject as EEG electrodes that had been positioned on the normal subjects to generate the database.

26. A system as in claim 25 wherein the means for detecting the analog magnetic fields includes a helmet-shaped dewar tail adapted to be positioned over the scalp of the subject, means for aligning the subject'"'"'s head relative to the dewar tail including fiducial marks on the dewar tail, and means or holding the subject'"'"'s head firmly in the dewar tail to prevent head movement.

27. A system as in claim 25 wherein the areas selected on the subject'"'"'s scalp correspond in location to areas to which at least some of the EEG electrodes of the 10/20 system are normally connected.

28. A system as in claim 25 and a SQUID as part of each magnetic field detector.

29. A system as in claim 25 and a magnetometer or gradiometer as part of each magnetic field detector.

30. A system as in claim 25 and means to Z transform or F ratio transform features extracted from the digital data derived from each magnetic field detector based upon the normative reference database.

31. A system as in claim 25 wherein the subject'"'"'s digital data is derived from 19 magnetic field detectors and is compared to data derived from 19 EEG electrodes corresponding in location to the full 10/20 system.

32. A system for detecting and analyzing the magnetic fields on the scalp of a subject arising from the subject'"'"'s brain activity, comprising:

(a) a plurality of magnetic field detectors arranged in an array in a dewar tail adapted to be positioned on the subject'"'"'s scalp, with the array corresponding in location to at least some of the EEG electrode locations in the 10/20 EEG system;

(b) means for amplifying and digitizing analog signals from the magnetic field detectors to produce a set of digital data corresponding to the analog signals from each magnetic field detector;

(c) computer means for automatically comparing quantitative features extracted from each set of digital MEG data taken under selected conditions or stimuli with features of Quantitative Electroencephalogram (QEEG) digital data collected from a normal group under the same conditions or stimuli and at the same scalp locations as the locations of the magnetic field detectors on the subject'"'"'s scalp.

33. A system as in claim 32 wherein the array comprises 19 magnetic field detectors arranged in locations corresponding to the locations of the 19 active electrodes of the 10/20 EEG system.

34. A system as in claim 32 wherein each magnetic field detector comprises a magnetometer or gradiometer and a SQUID and the dewar tail has a helmet-shaped cavity adapted to be brought into close proximity to the scalp of the subject.

35. A magnetoencephalogram (MEG) system for detecting and analyzing magnetic fields on the scalp of a subject generated by the subject'"'"'s brain activity, comprising:

(a) an array of a plurality of magnetic field detectors adapted to be positioned closely adjacent the subject'"'"'s scalp to detect and amplify magnetic fields from selected areas of the scalp;

(b) means for converting the detected magnetic fields into sets of digital data, with each set being generated by a magnetic field detector;

(c) computer means for automatically comparing quantitative features extracted from the sets of digital data with a normative reference database of such features simultaneously or previously collected from normal subjects, the comparisons being on a feature-by-feature basis, and the normal subjects having generated the database using EEG electrodes at the same relative locations on their scalps as the relative locations of the magnetic field detectors on the scalp of the subject;

(d) computer means for deriving a discriminant score based on the comparisons of (c) by selecting, weighting, and combining a number of said feature-by-feature comparisons;

(e) computer means for combining the discriminant scores to derive probabilities that the subject is classified in a specific diagnostic category;

(f) computer means for applying selected guardbands (rule-out levels) to the probabilities to classify the subject on the basis of the probabilities; and

(g) means for displaying the classification of the subject.

36. The system of claim 35 in which the normative reference database of features is used to generate age-corrected Z-scores which are in the same dimensional unit.

37. A magnetoencephalogram (MEG) system for analyzing the magnetic components of a human subject'"'"'s brain activity which are evoked by stimuli to determine the abnormality or normality of the response morphology, the system consisting of:

(a) a plurality of MEG magnetic-field detectors, i, a dewar tail containing the field detectors, means to bring the dewar tail proximate the scalp of the subject to detect the subject'"'"'s brain activity, the detectors being positioned in the same relative positions over the scalp as at least some of the electrodes are positioned in the 10/20 system;

stimuli means for automatically presenting a sequence of stimuli to the patient to evoke brain magnetic responses (MER) at the field detectors;

(b) amplifier means for amplifying the MER'"'"'s, A/D converter means for converting the amplified AMERs into digital data, averaging means for averaging the digital data to provide averaged magnetic evoked responses (AMERs) for analysis;

(c) computer means having a computer memory and calculation means for analyzing the AMERs by factor analysis, said computer means comparing each AMER from each field detector i to pre-formulated factor wave forms, j, stored in the computer memory to determine the power that each pre-formulated factor waveform, j, contributes to the AMER from detector i producing a factor score a_ij quantifying the contribution of each factor to the AMER from any field detector i;

(d) computer means for subjecting said factor score, a_ij, to Z-transform, Z_ij, or F-value, F_ij, automatically comparing each of said factor scores to the statistical parameters of a normative factor score database stored in the computer memory to produce a statistical assessment enabling a determination of response normality versus the degree of abnormality of morphology, and

(e) display means to display the said normality and degree of abnormality.

38. A system as in claim 37 wherein the display of (e) is a color coded topographic map comprising a head diagram in which each area of the head represents a magnetic detector location and the color represents a statistical estimate of the degree of abnormality reflected by the factor Z-score or F-value.

39. The system of claim 37 wherein the stimuli presented are a series of flashes or other visual stimuli to produce visual magnetic evoked responses (VMERs).

40. The system of claim 37 wherein the stimuli presented are a series of clicks or other auditory stimuli to produce auditory magnetic evoked potentials (AMERs).

41. The system of claim 37 wherein the stimuli presented are a series of electrical shocks or other somatosensory stimuli to produce somatosensory magnetic evoked responses (SMERs).

42. The system of claim 37 wherein a set of topographic maps depict the spatial (anatomical) distribution of each factor Z-score, the overall abnormality of the response morphology at each derivation $\sum _{}^{}$

i=iM

 



|Zi2

|1/2

and the residual variance unaccountable by the normal factors.