Real-time EEG spectral analyzer

• US 4,579,125 A
• Filed: 01/23/1984
• Issued: 04/01/1986
• Est. Priority Date: 01/23/1984
• Status: Expired due to Term

• Current Assignee: Aequitron Medical, Inc. (Medtronic Plc)
• Original Assignee: CNS Incorporated
• Inventors: Cohen, Daniel E., Strobl, Frederick T.
• Primary Examiner(s): Valentine, Donald R.
• Application Number:
  US06/572,868
• Time in Patent Office: 799 Days
• Field of Search: 128/731, 128/732, 128/733
• US Class Current: 600/544
• 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 5/742 using visual displays A61B5/339 takes precedence

Abstract

An electroencephelograph (EEG) signal analysis system automatically determines and displays, on a real-time basis, the frequency content of spontaneous EEG signals from the brain. The analog EEG signals from a plurality of channels are sampled and converted to digitized EEG waveforms during a continuing sequence of epochs of predetermined time duration. The digitized EEG waveform for each epoch and channel is transformed from the time domain to the frequency domain to produce a frequency spectrum representing amplitude of the EEG signal as a function of frequency. The digital data is screened both prior and subsequent to the transformation to identify those epochs which contain artifacts. During each update period, the frequency spectra for each channel from the artifact-free epochs are averaged. The amplitudes of each averaged frequency spectrum for frequencies in each of four basic frequency bands are then summed. The result is an amplitude value for each of the four frequency bands at each channel. The four amplitude values are used to update a display which contains a graphical representation of the amplitude in each of the four bands in a different color. The amplitude values are also compared with previously recorded base line values, and an audible or visual warning indication is provided when predetermined variances from the base line values occur.

Claims

1. A method of processing EEG signals from a plurality of channels to provide an output indicative of cerebral activity during a time period, the method comprising:

digitizing the EEG signals from the plurality of channels during a sequence of epochs of a known time duration to provide a digital waveform for each channel during each epoch;

transforming the digitized waveforms from a time domain to a frequency domain to produce a frequency spectrum for each epoch and channel;

averaging the frequency spectra for each channel over a plurality of epochs of the time period to produce an average frequency spectrum for each channel during the time period;

deriving, for each channel, frequency band amplitude values corresponding to a plurality of frequency bands based upon the average frequency spectrum for that channel; and

providing an output indicative of cerebral activity in each of the frequency bands at each channel during the time period based upon the frequency band amplitude values.

2. The method of claim 1 wherein the average frequency spectrum comprises an amplitude value for each of a plurality of frequencies, and wherein deriving frequency band amplitude values comprises summing the amplitude value of all frequencies of the average frequency spectrum within each frequency band.

3. The method of claim 2 wherein the plurality of frequency bands includes a delta EEG frequency band, a theta EEG frequency band, an alpha EEG frequency band, and a beta EEG frequency band.

4. The method of claim 1 and further comprising:

rejecting any epoch, prior to transforming the digitized waveforms, which includes a digitized waveform having a characteristic indicative of an artifact.

5. The method of claim 4 wherein the characteristic is a presence of amplitude in the digitized waveform which is greater than a threshold value.

6. The method of claim 4 wherein the characteristic is indicative of an eye movement artifact.

7. The method of claim 6 wherein the characteristic is a presence of amplitudes greater than a threshold value from anterior channels.

8. The method of claim 1 and further comprising:

rejecting any epoch, after transforming the digitized waveforms, which includes a frequency spectrum having a characteristic indicative of an artifact.

9. The method of claim 8 wherein the characteristic is indicative of a muscle artifact.

10. The method of claim 8 wherein the characteristic is a presence of high frequency components in the frequency spectra corresponding to temporal channels.

11. The method of claim 8 wherein the characteristics are indicative of a loose lead artifact.

12. The method of claim 1 wherein providing an output comprises:

providing a visual output in graphical form which includes a graph for each channel, each graph including a segment representative of each of the frequency bands, each segment having a relative size which is a function of the frequency band amplitude value for the frequency band corresponding to that segment.

13. The method of claim 12 wherein each segment has a size which is a function of a ratio of the frequency band amplitude value for that segment and a sum of the frequency band amplitude values for the channel.

14. The method of claim 13 wherein each graph has an area which is a function of the sum of the frequency band amplitude values for the channel.

15. The method of claim 14 and further comprising:

storing the frequency band amplitude value from a selected period as base line frequency band amplitude values; and

wherein the area of each graph is based upon a comparison of a sum of the frequency band amplitude values for the corresponding channel during the time period to a sum of the base line frequency band amplitude values for that channel.

16. The method of claim 12 wherein providing an output further comprises:

providing a graphical representation of a human subject'"'"'s head; and

positioning each of the graphs with respect to the graphical representation of the human subject'"'"'s head to indicate the respective locations of the channels corresponding to the graphs.

17. The method of claim 1 and further comprising:

storing the frequency band amplitude values from a selected time period as base line frequency band amplitude values;

comparing the frequency band amplitude values from each time period with the stored base line frequency band amplitude values; and

providing a warning indication when a variance between the frequency band amplitude values of the time period and the stored base line frequency band amplitude values exceeds a variance threshold.

18. The method of claim 17 wherein providing a warning indication includes:

providing an audible signal when the variance exceeds the variance threshold.

19. The method of claim 17 wherein providing the warning indication includes:

providing a visual indication of the channel which provided the frequency band amplitude values for which the variance exceeded the variance threshold.

20. The method of claim 1 wherein providing an output comprises:

providing a graph for each channel showing a series of time periods with a segment of variable size corresponding to each of the frequency band amplitude values for that channel during each time period of the series.

21. The method of claim 20 wherein the graph for each channel comprises a series of bar graphs arranged in side-by-side relationship and representing the series of time periods, and wherein the segment for each frequency band has a length which is a function of the frequency band amplitude value during a corresponding time period.

22. A method of processing an analog EEG signal to provide an output indicative of cerebral activity and providing an output which is periodically updated, the method comprising:

digitizing the analog EEG signal during a sequence of epochs of a known time duration to provide a digitized waveform for each epoch;

transforming the digitized waveform for each epoch from a time domain to a frequency domain to produce a frequency spectrum for that epoch;

averaging frequency spectra from a plurality of epochs corresponding to an update period to produce an average frequency spectrum;

deriving a delta EEG frequency band amplitude value, a theta EEG frequency band amplitude value, an alpha EEG frequency band amplitude value, and a beta EEG frequency band amplitude value based upon the average frequency spectrum; and

periodically updating a output to provide an indication of the cerebral activity in the delta, theta, alpha and beta EEG frequency bands during a completed update period based upon the frequency band amplitude values for that update period.

23. The method of claim 22 wherein the average frequency spectrum comprises an amplitude value for each of a plurality of frequencies, and wherein deriving frequency band amplitude values comprises summing the amplitude values of all frequencies of the average frequency spectrum within each frequency band.

24. The method of claim 22 and further comprising:

rejecting any epoch, prior to transforming the digitized waveform, for which the digitized waveform has a characteristic indicative of an artifact.

25. The method of claim 22 and further comprising:

rejecting any epoch, after transforming the digitized waveform, for which the frequency spectrum has a characteristic indicative of an artifact.

26. The method of claim 22 wherein periodically updating an output comprises:

providing a visual output in graphical form which includes a segment representative of each of the frequency bands, each segment having a relative size which is a function of the frequency band amplitude value for the frequency band corresponding to that segment.

27. The method of claim 26 wherein each segment has a size which is a function of a ratio of the frequency band amplitude value for that segment and a sum of the frequency band amplitude values.

28. The method of claim 27 wherein the graphical form has an area which is a function of the sum of the frequency band amplitude values.

29. The method of claim 28 and further comprising:

storing the frequency band amplitude values from a selected update period as base line frequency and amplitude values; and

wherein the area of the graphical form is based upon a comparison of a sum of the frequency band amplitude values during the update period to a sum of the base line frequency band amplitude values.

30. The method of claim 22 and further comprising:

storing the frequency band amplitude values from a selected update period as base line frequency band amplitude values;

comparing the frequency band amplitude values from each update period with the stored base line frequency band amplitude values; and

providing a warning indication when a threshold variance occurs between the frequency band amplitude values of the update period and the stored base line frequency band amplitude values.

31. The method of claim 22 wherein periodically updating an output comprises:

providing a graph showing a series of update periods with a segment of variable size corresponding to each of the frequency band amplitude values during each update period of the series.

32. A system for processing EEG signals from a plurality of channels to provide an output indicative of cerebral activity, the system comprising:

means for digitizing EEG signals from the plurality of channels during a sequence of epochs of a known time duration to provide a digital waveform of amplitude as a function of time for each channel during each epoch;

means for transforming the digitized waveforms for each epoch and channel from a time domain to a frequency domain to produce a frequency spectrum of amplitude as a function of frequency for each epoch and channel;

means for averaging the frequency spectra for each channel from a plurality of epochs of an update period to produce an average frequency spectrum for each channel during the update period;

means for deriving a frequency band amplitude value for each of a plurality of frequency bands based upon the average frequency spectrum for each channel;

means for displaying a visual representation of cerebral activity in each of the frequency bands at each channel as a function of display control signals; and

means for periodically updating the display control signals as each update period is completed based upon the frequency band amplitude values for that update period.

33. The system of claim 32 wherein the means for deriving a frequency band amplitude value sums the amplitude values of all frequencies of the average frequency spectrum within each frequency band to produce the frequency band amplitude value.

34. The system of claim 32 wherein the plurality of frequency bands includes a delta EEG frequency band, a theta EEG frequency band, an alpha EEG frequency band, and a beta EEG frequency band.

35. The system of claim 32 and further comprising:

means for rejecting any epoch which includes a digitized waveform having a characteristic indicative of an artifact.

36. The system of claim 32 and further comprising:

means for rejecting any epoch which includes a frequency spectrum having a characteristic indicative of an artifact.

37. The system of claim 32 wherein the visual representation comprises a graph for each channel, each graph including a segment representative of each of the frequency bands, each segment having a relative size which is a function of the frequency band amplitude value for the frequency band corresponding to that segment.

38. The system of claim 37 wherein each each segment has a a size which is a function of a ratio of the frequency band amplitude value for that segment and a sum of the frequency band amplitude values for the channel.

39. The system of claim 38 wherein the means for periodically updating determines an area of each graph as function of the sum of the frequency band amplitude values for the channel corresponding to that graph.

40. The system of claim 39 and further comprising:

means for storing the frequency band amplitude values from a selected update period as base line frequency band amplitude values; and

wherein the means for periodically updating determines the area of each graph based upon a comparison of a sum of the frequency band amplitude value for the corresponding channel during the update period to a sum of the base line frequency band amplitude value for that channel.

41. The system of claim 37 wherein the visual representation further comprises:

a graphical representation of a human subject'"'"'s head, the graphs being positioned with respect to the graphical representation of the human subject'"'"'s head to indicate the respective locations of the channels corresponding to the graphs.

42. The system of claim 32 and further comprising:

means for storing the frequency band amplitude values from a selected update period as base line frequency band amplitude values;

means for comparing the frequency band amplitude values from each update period with the stored base line frequency band amplitude values; and

means for providing a warning indication when a threshold variance occurs between the frequency band amplitude values of the update period and the stored base line frequency band amplitude values.

43. The system of claim 32 wherein the visual representation comprises a graph for each channel showing a series of update periods with a segment of variable size corresponding to each of the frequency band amplitude values for that channel during each update period of the series.