Method and system for three dimensional tomography of activity and connectivity of brain and heart electromagnetic waves generators

• US 5,307,807 A
• Filed: 03/11/1992
• Issued: 05/03/1994
• Est. Priority Date: 03/15/1991
• Status: Expired due to Term

• Current Assignee: Centro de Neurociencias de Cuba (CNEURO)
• Original Assignee: Centro de Neurociencias de Cuba (CNEURO)
• Inventors: Gonzalez Andino, Sara L., Biscay Lirio, Rolando J., de Peralta Menendez, Rolando G., Riera Diaz, Jorge J., Bayard, Jorge B., Valdes Sosa, Pedro A., Pascual Marqui, Roberto D.
• Primary Examiner(s): Cohen, Lee S.
• Assistant Examiner(s): Casler, Brian L.
• Application Number:
  US07/849,594
• Time in Patent Office: 783 Days
• Field of Search: 128/653.1, 128/731, 606/130, 324/248, 324/244, 324/245, 324/260
• US Class Current: 600/409
• CPC Class Codes: A61B 5/246 using evoked responses
• CPC Class Codes: A61B 5/377 using evoked responses

Abstract

A method and system for the localization and characterization of the generators of human brain electromagnetic physiological activity includes a set of bioelectromagnetic amplifiers, sensorial stimulators, and a computer based system for signal analog to digital conversion and recording. Sufficient statistics, including higher order statistical moments, for event related components are computed from the recorded signals, either in the time, frequency, or time-frequency domain, retaining stationary, non-stationary, linear, and non-linear information. The localizations, orientations, activities, and connectivities of the generators are obtained by solving the inverse problem using sufficient statistics under anatomical and functional constraints. Realistic head geometry and conductivity profiles are used to transform the measurements into infinite homogeneous medium measurements, through use of ananatomical deconvolution operator, thus simplifying optimally inverse solution computations. Goodness of fit tests for the inverse solution are provided. Generator characteristics are visually displayed in the form of three and two dimensional head images, and optionally include probability scaled images obtained by comparing estimated generator characteristics with those of a normal population sampled and stored in a normative data base.

Claims

1. A method for three dimensional tomography of activity and connectivity of brain electromagnetic waves generators, said method including the steps of:

a) positioning a set of electrodes and magnetic sensors to detect and record brain electromagnetic signals representing physiological activity in the form of an electroencephalogram (EEG) and a magnetoencephalogram (MEG), and measuring exact positions of the electrodes and sensors with respect to a reference coordinate system determined by certain anatomical landmarks of a head of an experimental subject;

b) amplifying the electromagnetic signals detected at each electrode and sensor by connecting an amplifier to each electrode and sensor;

c) obtaining on-line digital spatio-temporal signals, consisting of said EEG and MEG, by connecting analog-digital converters to each amplifier, and digitizing all data as it is gathered;

d) determining a parametric description for an anatomy of the experimental subject'"'"'s head to obtain a descriptive parametric geometry;

e) using said descriptive parametric geometry for constructing a head phantom with volume conductor properties of the experimental subject'"'"'s head;

f) performing EEG and MEG measurements on said head phantom due to known current dipoles located in a corresponding neural tissue volume, for determining a linear operator which transforms original EEG and MEG measurements into equivalent infinite homogeneous medium measurements (anatomical deconvolution);

g) using said descriptive parametric geometry for determining anatomical and functional constraints for localizations, orientations, activities, and connectivities of the brain electromagnetic waves generators (generator constraints);

h) digitally pre-processing the EEG and MEG for artifact and noise elimination, and for separation of EEG and MEG samples related to said fiducial markers, for obtaining event related components (ERCs);

i) statistically analyzing said ERCs for determining the most adequate numerical description of spatio-temporal properties in terms of sufficient statistics;

j) computing activities and connectivities of the ERCs generators, based on a static solution to an inverse electromagnetic problem, under said generator constraints, using sufficient statistics for the ERCs transformed to an infinite homogeneous medium by means of said anatomical deconvolution;

k) in case said generator constraints do not allow a unique solution to the inverse problem, decreasing a number of ERCs generators sufficiently to allow for proper identifiability of the inverse problem;

l) statistically evaluating a goodness of fit of the static solution, taking into account an existence of colored spatial and temporal noise, and including statistical hypotheses for testing an absence of activity and connectivity of the ERCs generators; and

m) visually displaying three dimensional and two dimensional images corresponding to the localizations, orientations, activities, and connectivities of the ERCs generators.

2. A method as claimed in claim 1 wherein events are detected in the EEG and MEG for recording and modification of an experimental sequence.

3. A method as claimed in claim 2 wherein said events are defined deterministically.

4. A method as claimed in claim 2 wherein said events are defined statistically.

5. A method as claimed in claim 2 wherein said events are determined in terms of fuzzy logic.

6. A method as claimed in claim 2 wherein said events are paroxistic events.

7. A method as claimed in claim 2 wherein said events are alpharhythmic desynchronization events.

8. A method as claimed in claim 1 wherein series expansions are established in terms of three dimensional basis functions for the description of contours, physical properties, and metabolic properties of the head, producing a descriptive parametric geometry consisting of sets of expansion coefficients for the basis function.

9. A method as claimed in claim 1 wherein the descriptive parametric geometry establishes a priori constraints on said localizations and orientations of the ERCs generators.

10. A method as claimed in claim 1 wherein the descriptive parametric geometry is used for constructing the head phantom with all of said volume conductor properties of the experimental subject'"'"'s head.

11. A method as claimed in claim 1 wherein EEG and MEG measurements are performed on the head phantom, using established current dipoles located in the corresponding neural tissue volume, with known locations, orientations and activities, where theoretical EEG and MEG values are computed in the corresponding infinite homogeneous medium, and where the linear operator in matrix form which transforms original EEG and MEG measurements into equivalent infinite homogeneous medium measurements is computed in a least squares sense, thus defining an anatomical deconvolution operation.

12. A method as claimed in claim 1 wherein event related samples (ERSs) of said EEG and MEG are separated according to different types of fiducial markers, and wherein said ERSs contain ERCs, which are estimated under the assumption that a set of generators of the ERCs of said ERSs have fixed localizations and orientations, and stochastic activities and connectivities.

13. A method as claimed in claim 1 wherein statistical hypothesis tests are used for determining a set of sufficient statistics for the description of the ERCs, where said set of sufficient statistics consist of cumulants of any order, in any domain.

14. A method as claimed in claim 1 wherein the ERSs are modeled as containing additive noise due to the activity of diffuse generators independent of ERCs concentrated generators, said diffuse generators having the properties of a stochastic process.

15. A method as claimed in claim 1 wherein the inverse problem is solved for the activities and connectivities of neuronal generators, based on all the information provided by ERCs sufficient statistics, under the generator constraints on localizations and orientations provided by the descriptive parametric geometry, for infinite homogeneous medium measurements obtained by applying an anatomical deconvolution operator.

16. A method as claimed in claim 1 wherein the goodness of fit of the estimated inverse solution and tests of hypotheses on no activity of the ERCs generators examined based on statistical resampling techniques.

17. A method as claimed in claim 1 wherein the estimated localizations, orientations, activities, and connectivities of the ERCs generators of an experimental subject are compared with those of a normal population by means of multivariate metrics for measuring distances between estimators and normative data of a sample from the normal population, taking into account an effect of convariables in order to decrease metric variability.

18. A method as claimed in claim 1 wherein visual displays are presented in the form of three dimensional and two dimensional images of the head, where the localizations, orientations, activities, and connectivities of the ERCs generators are displayed by coding their numerical values in terms of color, intensity, and graphical icons.

20. A method as claimed in claim 1 further comprising the step of pre visual, auditory, and somato-sensorial stimulation to the experimental subject during said recording of the EEG and MEG, carried out under control of said central experimental program.

21. A method as claimed in claim 20 further comprising the step of identifying responses produced by the experimental subject during said EEG and MEG recording, inclusion of fiducial markers of the type and time of stimulation in said recording, and modifying an experimental program sequence based on the responses produced by the experimental subject.

22. A method as claimed in claim 21 further comprising the step of including fiducial markers in said recording for the time and type of spontaneous events detected in the EEG and MEG produced by the experimental subject during recording, and modifying the central experimental program based on the spontaneous events.

23. A method as claimed in claim 1 wherein step d) comprises the step of determining the parametric geometry for the anatomy of the experimental subject'"'"'s head by means of exact computations based on image processing of the subject'"'"'s head.

24. A method as claimed in claim 1 wherein step d) comprises the step of determining the parametric description for the anatomy of the experimental subject'"'"'s head based on approximate computations based on a small set of anatomical measurements and comparison with a data base of normal and abnormal variability.

25. A method as claimed in claim 1 further comprising the step of computing multivariate distances between ERC generators characteristic of said experimental subject and of a normal population as determined by a normative data-base.

26. A method as claimed in claim 25 further comprising the step of displaying said multivariate distances.

27. A method as claimed in claim 1 further comprising the step of presenting stimulation sequences to the subject in the form of computerized video-games, and recording responses to the video-games.

28. A method as claimed in claim 1 wherein the descriptive parametric geometry is obtained from functional images of the experimental subject'"'"'s head.

29. A method as claimed in claim 1 wherein the descriptive parametric geometry is obtained from a minimum set of easily measured parameters which can be used to statistically predict the descriptive parametric geometry via information contained in a normative data base with a representative sample covering a typical population.

30. A method as claimed in claim 1 wherein EEG and MEG values are theoretically computed on a head phantom defined by the descriptive parametric geometry using established current dipoles located in corresponding neural tissue volume, with known locations, orientations, and activities, and wherein the theoretical EEG and MEG values are computed in the corresponding infinite homogeneous medium, and wherein a linear operator in matrix form which transforms original EEG and MEG measurements into equivalent infinite homogeneous medium measurements is computed in a least squares sense, thus defining an anatomical deconvolution operation.

31. A method as claimed in claim 1 wherein statistical hypothesis tests are used for determining a set of sufficient statistics for a description of stochastic event related samples (ERSs), wherein said set of sufficient statistics consists of multiple time series parametric models, in a domain selected from the group consisting of the time, frequency, and time-frequency domains.

19. A method as claimed in 18 wherein multivariate metrics corresponding to comparison with norms are also displayed by superposition.

32. A system for the three dimensional tomography of activity and connectivity of brain electromagnetic waves generators comprising:

a) A set of electrodes and magnetic sensors adapted to be positioned to detect brain electromagnetic signals representative of physiological activity in the form of an electroencephalogram (EEG) and a magnetoencephalogram (MEG), and means for measuring exact positions of the electrodes and sensors with respect to a reference coordinate system determined by certain anatomical landmarks of a head of an experimental subject.

b) Means for amplification of said electromagnetic signals detected at each electrode and sensor;

c) Means for obtaining on-line digital spatio-temporal signals consisting of said EEG and MEG;

d) Means for the presentation of visual, auditory, and somato-sensorial stimulation to the experimental subject during EEG and MEG recording;

e) Means for recording vocal or movement responses produced by the experimental subject during EEG and MEG recording;

f) A central digital computer subsystem, comprising;

means for reading the experimental subject'"'"'s image data, and means for computing and storing a descriptive parametric geometry, an anatomical deconvolution operator, and generator constraints;

means for constructing a head phantom based on the descriptive parametric geometry, and means for the implantation of current dipoles in a corresponding neural tissue volume of the phantom;

means controlling experiments including means for causing stimulation of the experimental subject, recording of the subject'"'"'s responses, detection and recording of special EEG and MEG events, and simultaneous recording of said electromagnetic signals;

means for pre-processing the recorded electromagnetic signals for artifact and noise elimination;

means for estimating event related components (ERCs)

means for computing sufficient statistics of the ERCs;

means for estimating an additive non-white spatio-temporal noise due to diffuse generators;

means for performing tests of hypothesis about the goodness of fit of an estimated inverse solution;

means for estimating localizations, orientations, activities, and connectivities of generators of the ERCs;

means for comparing characteristics of the ERCs generators with a normative data base and means for computing multivariate metrics;

means for the visual display of ERCs generators characteristics and of the multivariate metrics.

33. A system as claimed in claim 32 wherein said central digital computer subsystem includes a multitasking processor.

34. A system as claimed in claim 32 wherein said central digital computer subsystem includes a set of distributed processors.

35. A system as claimed in claim 32 wherein said means for reading the experimental subject'"'"'s image data comprises means for reading images selected from the group consisting of CAT scan images, NMR images, PET images, and anatomical measurements.