Method and apparatus for multifaceted electroencephalographic response analysis (MERA)

• US 5,363,858 A
• Filed: 05/05/1993
• Issued: 11/15/1994
• Est. Priority Date: 02/11/1993
• Status: Expired due to Term

• Current Assignee: Brainwave Science Incorporated, Central Intelligence Agency, Defense Counterintelligence and Security Agency, Forensic Science Laboratory Delhi, Francis Luca Conte, Prime Technology Incorporated (General Dynamics Corporation), Government of New Zealand, Government of the United States of America, Harvard University, Islamic Republic of Pakistan, Kingdom of Saudi Arabia, National Center For Credibility Assessment, Republic of India, The Republic of South Africa, Ukraine, United Kingdom of Great Britain and Northern Ireland, Federal Bureau of Investigation (United States Department of Justice), University of Canterbury, University of Illinois at Urbana Champaign (University of Illinois), University of Washington, United States Navy (U.S. Department Of Defense)
• Original Assignee: Francis Luca Conte
• Inventors: Farwell, Lawrence A.
• Primary Examiner(s): Kamm, William E.
• Assistant Examiner(s): Getzow, Scott M.
• Application Number:
  US08/057,607
• Time in Patent Office: 559 Days
• Field of Search: 128/731, 128/732, 128/639, 607/139, 607/140
• US Class Current: 600/544
• CPC Class Codes: A61B 5/163 by tracking eye movement, gaze, or pupil change
• CPC Class Codes: A61B 5/164 Lie detection
• CPC Class Codes: A61B 5/291 for electroencephalography [EEG]
• CPC Class Codes: A61B 5/374 Detecting the frequency distribution of signals, e.g. detecting delta, theta, alpha, beta or gamma waves
• CPC Class Codes: A61B 5/377 using evoked responses
• CPC Class Codes: A61B 5/7207 of noise induced by motion artifacts

Abstract

A method and apparatus for multifaceted electroencephalographic response analysis (MERA) is described. MERA involves presenting stimuli to a subject through at least one of the visual and auditory modalities, recording electrophysiological brain responses indicative of information processing in response to said stimuli, and analyzing multiple features of said brain responses. This technique is used both to detect concealed information in the brain and to communicate directly from a brain to a computer, and thus to command computers and also to command mechanical devices. The memory and encoding related multifaceted electroencephalographic response (MERMER), a particular brain response elicited by, and unfolding within two seconds of the onset of, noteworthy stimuli, involving an electrically positive aspect generally maximal parietally followed by an electrically negative aspect with substantial frontal amplitude, and also involving characteristic changes in the frequency domain, is detected, analyzed, and applied for the above stated purposes. The MERMER is elicited by stimuli that refer to concealed information noteworthy to an individual and also by stimuli that are noteworthy due to a subject'"'"'s desire to communicate the information to which they refer. Responses are compared with at least one of a) responses to other stimuli known to be irrelevant, and b) responses to other stimuli known to be noteworthy, in order to determine the presence or absence of the MERMER in response to the stimuli in question. In this manner, concealed information is detected, and information of significance is communicated.

Claims

1. A method of detecting concealed information stored in the human brain through multifaceted electroencephalographic response anaysis (MERA) comprising the steps of:

1) Presenting stimuli to a subject;

2) Recording electrical brain responses to the stimuli;

3) Analyzing said electrical brain responses for uncovering as a first facet an electrically positive P3 component thereof, and as a second facet at least one of a phasic shift in frequency power spectrum thereof, and an electrically negative component thereof following in time said P3 component.

2. The method in claim 1 wherein data signals corresponding to said electrical brain responses to different stimulus types are compared, and the stimuli include:

1) Irrelevant stimuli ("irrelevants");

2) Stimuli relevant to the information to be detected, that are not identified in instructions to the subject as being different from the irrelevants, and would not be clearly distinguishable from the irrelevants for a subject lacking the specified concealed information being sought ("probes");

3) At least one of the following ("targets");

a) Stimuli that are identified to the subject as being noteworthy, and in response to which the subject is instructed to perform a task, that are relevant to the information being detected; and

b) Stimuli that are identified to the subject as being noteworthy, and in response to which the subject is instructed to perform a task, that are NOT relevant to the information being detected.

3. The method in claim 2 wherein the data signals that are collected and analyzed include data signals from brain electrical activity from the time period of 0 to 1400 msec post-stimulus and data signals from brain electrical activity from the time period of 1400 to 2000 msec post-stimulus.

4. The method in claim 3 wherein the data signals include signals corresponding to a widespread late negative component with a substantial frontal amplitude ("late negative component"), and wherein said frontal data signals contribute materially to the analysis.

5. The method in claim 3 wherein the analysis includes bootstrapping across single trails.

6. The method in claim 2 wherein the data signals are collected from more than one scalp location, and a computer-implemented signal processing and analysis method is applied that includes analysis of more than one scalp location.

7. The method in claim 6 wherein the data signals that are collected and analyzed include some data signals from the time period of 0 to 1400 msec post-stimulus and some data signals from the time period of 1400 to 2000 msec post-stimulus.

8. The method in claim 7 wherein the analysis method includes determining

a subsequent negative component with a widespread scalp distribution including a substantial frontal amplitude.

9. The method in claim 8 wherein the analysis includes computation of the correlations between the probe and irrelevant responses and between the probe and target responses.

10. The method in claim 9 wherein the analysis includes bootstrapping across single trails.

11. The method in claim 8 wherein the analysis includes bootstrapping across single trails.

12. The method in claim 7 wherein the analysis includes bootstrapping across single trails.

13. The method in claim 9 wherein the analysis method includes determining:

1) Base-to-peak amplitude of the P3 and late negative components, at the respective scalp locations where each is maximal;

2) Peak-to-peak amplitude of the P3 and late negative components, at the respective scalp locations where each is maximal;

3) Absolute value area of the P3 and late negative components, at the respective scalp locations where each is maximal;

4) Analysis epochs that can be adjusted, based on factors including onset and offset of the components of interest; and

5) Pairwise correlation of responses to the stimulus types to detect similarity between the probe response and at least one of the target and irrelevant responses; and

the analysis method further includes;

6) Bootstrapping across single trials;

7) Combining the results obtained with at least two different comparison metrics;

8) Determining whether or not the information signaled by the probe stimuli has been previously stored in the brain of the subject;

9) Obtaining a statistical confidence level for said determination of step (8); and

10) Collecting data from multiple electroencephalographic channels and at least one channel to record eye movement artifacts.

14. The method in claim 13 wherein frequency domain results are included in the analysis.

15. The method in claim 6 wherein the analysis includes bootstrapping across single trials.

16. The method in claim 2 wherein the analysis includes at least one of the following:

1) Transformation of the data signals to the frequency domain and comparison of the probe frequency spectrum with at least one of the target and irrelevant frequency spectra;

2) Computation of time frequency data for at least one of the probe, target, and irrelevant responses.

17. The method in claim 16 wherein the differences between pre-stimulus and post-stimulus frequency spectra are computed and compared for the responses to the probe, target, and irrelevant stimuli.

18. The method in claim 16 wherein time frequency data are computed and compared for the responses to the probe, target, and irrelevant stimuli.

19. The method in claim 16 wherein the frequency domain responses to the probe, target, and irrelevant stimuli are compared using bootstrapping.

20. The method in claim 2 wherein precisely timed visual and auditory stimuli are presented simultaneously.

21. A method of applying multifaceted electroencephalographic response analysis (MERA) comprising:

1) Making conscious choices by a subject from among at least two options;

2) Measuring electrical brain activity of said subject in the form of electrical signals that is associated with the choice of options;

3) Analyzing the electrical signals obtained for uncovering as a first facet an electrically positive P3 component thereof, and as a second facet at least one of a phasic shift in frequency power spectrum thereof, and an electrically negative component thereof following in time said P3 component to detect which option is chosen by said subject; and

4) Utilizing the information obtained regarding the options chosen to perform a function using a device actuated in response to said chosen option.

22. The method in claim 21 wherein said device is a computer.

23. The method in claim 22 wherein said computer controls a mechanical device having motion capability.

24. The method in claim 23 wherein electrical signals are analyzed from more than one active scalp location.

25. The method in claim 23 wherein the signal analysis includes both analysis in the time domain and analysis in the frequency domain.

26. The method in claim 23 wherein the signals measured are event-related brain potentials including a P3 component.

27. The method in claim 23 wherein the analysis includes at least one of the following:

correlation and covariance with a template.

28. The method in claim 23 wherein said device is a robot having motion capability.

29. The method in claim 23 wherein the options chosen among are presented in at least one of the following two modes:

1) visually presented on a video screen under computer control, and

2) presented through the auditory modality.

30. The method in claim 22 wherein the options chosen among are presented to the subject by said computer.

31. The method in claim 23 wherein the options chosen among are presented in at least one of the following two modes:

1) visually presented on a video screen under computer control, and

2) presented through the auditory modality.

32. The method in claim 22 wherein said computer controls a speaker for communication of verbal material.

33. The method in claim 22 wherein the signal analysis includes both analysis in the time domain and analysis in the frequency domain.

34. The method in claim 22 wherein the electrical signals are measured over a time epoch and the analysis includes at least some of the time epoch between the onset of a stimulus and 600 msec post-stimulus, and at least some of the time epoch between 600 msec post-stimulus and 2000 msec post-stimulus.

35. The method in claim 34 wherein the time epoch analyzed includes at least some of the period between 1000 msec post-stimulus and 2000 msec post-stimulus.

36. The method in claim 22 wherein both positive and negative deflections in the voltage potential of the measured electrical signals contribute materially to the signal analysis.

37. The method in claim 22 wherein electrical signals are analyzed from more than one active scalp location.

38. A method of analyzing electrical brain activity consisting of:

1) Presenting stimuli to a subject;

2) Detecting electrophysiological manifestations of information-processing brain activity in response to said stimuli;

3) Amplifying, digitizing, and analyzing said manifestations of information-processing brain activity;

4) Applying a computer-implemented signal analysis method that includes analysis in both the time and frequency domains, and mathematically combines the results of said analysis in said two domains to make inferences regarding said information-processing brain activity.

39. The method in claim 38 wherein the brain response analyzed contains a memory and encoding related multifaceted electroencephalographic response (MERMER).

40. The method in claim 39 wherein the MERMER is used for at least one of the following:

1) Detection of concealed information;

2) Communication between a brain and a computer;

3) Communication of verbal material from a brain to a computer, and from the computer to another person through a device that generates verbal output through at least one of the visual and auditory modalities;

4) Communication from a brain to a computer of commands for the control of a mechanical device; and

5) Communication from a brain to a computer of commands for the control of a robot.

41. A headband for measurement of electrical brain activity from a scalp of a subject, said scalp having frontal, central, parietal, left mastoid, and right mastoid locations, said headband comprising:

a flexible band having first and second opposite ends and a center portion therebetween, and being sized to surround said scalp for joining together said first and second ends;

a flexible overflap having first and second opposite ends, with said overflap first end being integrally joined to said band center portion, and being sized to extend over said scalp frontal, central, and parietal locations for joining together said overflap second end to said band first and second ends;

first means for releasably joining together said band first and second ends;

second means for releasably joining together said overflap second end with said joined together band first and second ends;

a plurality of first pads each having an electrode wire fixedly joined thereto, with an electrode snap joined to a distal end of said wire for mounting an electrode thereto; and

third means for releasably joining said first pads to said overflap for allowing position adjustment of said electrode snaps along said overflap, and thereby position adjustment of said electrode snaps relative to said scalp.

42. A headband according to claim 41 wherein said electrode wires extend from said first pads and through said overflap, and have proximate ends joined to a connector disposed adjacent to said overflap second end.

43. A headband according to claim 41 wherein said third joining means includes a primary Velcro pad fixedly joined to said overflap, and said first pads are in the form of cooperating secondary Velcro pads for allowing each of said first pads to be releasably joined to said primary Velcro pad.

44. A headband according to claim 43 wherein said primary pad is elongate and longer than said secondary pad for allowing position adjustment of said secondary pad along said primary pad.

45. A headband according to claim 44 further comprising:

a plurality of second pads each having an electrode wire fixedly joined thereto, with an electrode snap joined to a distal end thereof for mounting an electrode thereto; and

fourth means for releasably joining said second pads to said band for allowing position adjustment of said second pad electrode snaps along said band.

46. A method of detecting information stored in a brain of a subject comprising:

presenting to said subject relevant and irrelevant stimuli;

detecting an electrical brain response from said subject in response to each of said stimuli;

analyzing said electrical brain responses for uncovering as a first facet an electrically positive P3 component thereof, and as a second facet at least one of a phasic shift in frequency power spectrum thereof and an electrically negative component thereof following in time said P3 component; and

comparing said electrical brain responses due to said relevant and irrelevant stimuli for detecting said information in said subject brain based on said second facet.

47. A method according to claim 46 wherein said detecting step for each of said stimuli has a predetermined time period selected for initially detecting said first facet and then detecting said second facet.

48. A method according to claim 47 wherein:

said subject has a scalp with frontal, central, and parietal locations;

said detecting step is effective for detecting said electrical brain responses at said frontal, central, and parietal scalp locations; and

said electrical brain responses are characterized by being largest at different ones of said scalp locations for said first and second facets.

49. A method according to claim 47 wherein said second facet includes a frequency power spectrum characterized by power first increasing, decreasing, and then increasing as frequency increases.

50. A method according to claim 47 wherein said detecting step is effective for detecting said electrical brain responses at frequencies below 6 Hz.

51. A method according to claim 50 wherein said detecting step is effective for detecting said electrical brain responses at frequencies between about 0.1 and 2.0 Hz.

52. A method according to claim 47 wherein said stimuli further comprise:

a Probe stimulus representing relevant specific information to be detected in said subject brain, and effecting said second facet in a subject having knowledge thereof, and not effecting said second facet in a subject not having knowledge thereof;

a Target stimulus representing information for which said subject is required to perform a predetermined response and which effects said second facet in said subject; and

an Irrelevant stimulus not relevant to said specific information to be detected in said subject brain and not effecting said second facet in said subject.

53. A method according to claim 52 wherein said comparing step further comprises obtaining a difference in said electrical brain responses for said Probe and Irrelevant stimuli, and obtaining a difference in said electrical brain responses for said Target and Irrelevant stimuli.

54. A method according to claim 52 wherein said comparing step further comprises obtaining a normalized waveform by subtracting an average waveform of said electrical brain responses for said Target and Irrelevant stimuli from a waveform for said electrical brain responses of said Probe stimulus.

55. A method according to claim 52 wherein said normalized waveform has one shape and distribution when said relevant specific information is detected in said subject, and a generally opposite shape and distribution when said relevant specific information is not detected in said subject.

56. A method according to claim 47 further comprising performing a function using a device actuated in response to detection of said information in said subject.

57. A method of detecting information stored in a brain of a subject comprising:

presenting to said subject relevant and irrelevant stimuli;

detecting an electrical brain response from said subject in response to each of said stimuli;

analyzing said electrical brain responses for uncovering an electrically negative component thereof having an onset latency greater than about 800 msec and having a peak thereafter; and

comparing said electrical brain responses due to said relevant and irrelevant stimuli for detecting said information in said subject brain based on said negative component.