Processing EEG signals to predict brain damage

• US 6,931,274 B2
• Filed: 05/06/2002
• Issued: 08/16/2005
• Est. Priority Date: 09/23/1997
• Status: Expired due to Fees

• Current Assignee: Natus Medical Incorporated
• Original Assignee: Tru-Test Limited
• Inventors: Williams, Christopher Edward
• Primary Examiner(s): Nasser, Robert L
• Application Number:
  US10/140,360
• Publication Number:
  US 20030023183A1
• Time in Patent Office: 1,198 Days
• Field of Search: 600/300, 600/544, 600/545
• 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 5/4094 Diagnosing or monitoring seizure diseases, e.g. epilepsy
• CPC Class Codes: A61B 5/412 Detecting or monitoring sepsis
• CPC Class Codes: A61B 5/4806 Sleep evaluation A61B5/4821 takes precedence; devices for inducing sleep A61M21/02
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal
• CPC Class Codes: A61B 5/721 using a separate sensor to detect motion or using motion information derived from signals other than the physiological signal to be measured
• CPC Class Codes: A61B 5/7232 involving compression of the physiological signal, e.g. to extend the signal recording period
• CPC Class Codes: A61B 5/7257 using Fourier transforms
• CPC Class Codes: A61B 5/726 using Wavelet transforms

Abstract

Rapid and accurate in-vivo assessment of cerebral white matter injury particularly for pre-term infants, for timely treatment and/or prediction of outcomes has been very limited. This invention exploits the discovery that reduced high-frequency EEG intensity, particularly as shown by the upper spectral edge frequency, is a good indicator of cerebral white matter neural injury and is well correlated with MRI results. With more experience of clinical cases, a set of simple rules such as “if the spectral edge value is below 8 Hz there is a high likelihood of injury” may be validated, yet the EEG technology involved is largely invisible to the user. In the invention, EEG signals are processed by software to obtain, store, and graphically display bilaterally collected EEG spectral edge and intensity values over from hours to weeks. Rejection of corrupted signals by filtering and gating means is responsive to incoming signal characteristics, to additional inputs such as motion sensors or impedance tests, and to patient data (gestational age in particular). The invention includes the software and methods of use.

Claims

1. An apparatus for processing an electroencephalogram (EEG) signal from a subject to assist in management of possible cerebral white-matter neural injury in the subject, comprising:

(a) apparatus to acquire at least one channel of an EEG signal from the subject over a period of time, and

(b) computing apparatus programmed to analyze the frequency distribution of the EEG signal so acquired, from within a range of from 1 Hz to 50 Hz, and to produce condensed output information descriptive and/or predictive of cerebral white-matter neural injury in the subject.

2. The apparatus of claim 1 further comprising apparatus programmed to analyze the frequency distribution by determining a numerical value for the upper spectral edge of the frequency distribution of the intensity of the EEG signals.

3. The apparatus of claim 2 further comprising apparatus to determine the spectral edge for each of a series of time intervals, to store a corresponding series of spectral edge values, and to include the series of spectral edge values on a display, thereby to enable forecasting of an outcome of cerebral white-matter neural injury in the subject.

4. The apparatus of claim 3 further comprising apparatus to determine, within the range of 1 Hz to 35 Hz, the upper spectral edge below which about 95% of the EEG intensity occurs.

5. The apparatus of claim 4 further comprising apparatus to provide a comparison of the determined upper spectral edge with stored EEG spectral edge and associated neurological outcome information from previous cases, thereby to enable forecasting of an outcome of cerebral white-matter neural injury in the subject.

6. The apparatus of claim 5 further comprising apparatus to acquire a plurality of channels of EEG signals, so that spectral edge recordings are made from a plurality of positions on the head of the subject, thereby to improve confidence in forecasting outcome of cerebral white-matter neural injury in the subject.

7. A program embodied on a computer readable medium in combination with the apparatus of claim 5, the program comprising a routine to display comparable portions of both a current record and a previously collected record thereby permitting a comparison of the current record with stored information derived from the EEG and neurological outcome information from previous cases, thereby enabling prediction of an outcome of cerebral white-matter neural injury in the subject.

8. The apparatus of claim 4 further comprising apparatus programmed to eliminate an artifact in output information by action of at least one event-responsive filtering procedure to delete from the output information data likely to be in error.

9. The apparatus of claim 8 where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of a condition within the EEG signal selected from abnormally low signal amplitude, abnormally high signal amplitude, a signal indicating a seizure in the subject, a signal indicative of sleep in the subject, and a signal indicative of a presence of electrical interference.

10. The apparatus of claim 8 further comprising a movement sensor and where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of movement.

11. The apparatus of claim 8 further comprising an electrode impedance detector and where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edged during detection of an incorrect electrode impedance.

12. The apparatus of claim 8 further comprising an electrode identification detector and where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of an inappropriate electrode.

13. The apparatus of claim 8 further comprising an input apparatus through which an operator can input information describing the developmental status of the subject and apparatus programmed to make alterations to filter parameters in accordance with known neurophysiological parameters of subjects of that developmental status.

14. The apparatus of claim 8 further comprising apparatus for issuing an appropriate advisory message to an operator for the operator to cause removal of the artifact.

15. A program embodied on a computer readable medium in combination with the apparatus of claim 8, the program comprising a routine to display an annotated image of a head to indicate a suitable position for the placement of EEG electrodes.

16. A program embodied on a computer readable medium in combination with the apparatus of claim 8, the program comprising at least one routine to accept data representing a digitized EEG signal and filter the data to exclude data likely to be in error, thereby facilitating prediction of the outcome of cerebral white-matter neural injury in the subject.

17. The program embodied on a computer readable medium of claim 16 further comprising at least one routine to receive information and to act on the data in response to the information to minimize the effect of an artifact.

18. The program embodied on a computer readable medium of claim 17 further comprising at least one routine to analyze the incoming data to establish the likely presence of an artifact, and, if the likely presence of an artifact is established, to output an advisory message to an operator to cause the removal of the artifact.

19. The apparatus of claim 3 further comprising apparatus to determine, between 2 Hz and 20 Hz, the upper spectral edge below which about 90% of the EEG intensity occurs.

20. A method for using the apparatus of claim 3 including screening at least one subject from time to time by using the apparatus to define the characteristics of the EEG of the subject to determine whether the subject being screened is at risk of cerebral white-matter neural injury to allow treatment and or forecasts of likely outcomes to be made accordingly.

21. A method for using the apparatus of claim 3 including continuously monitoring at least one subject by using the apparatus to define and follow the characteristics of the EEG of the subject in order to determine the presence and extent of cerebral white-matter neural injury to allow effectiveness of any treatment of the subject to be monitored and/or to allow a forecast of likely outcome to be made.

22. A program embodied on a computer readable medium in combination with the apparatus of claim 2, the program comprising a routine to accept data representing a digitized EEG signal and provide an output representing a numerical value for the upper spectral edge of the frequency distribution of the intensity within the EEG signal.

23. A program embodied on a computer readable medium in combination with the apparatus of claim 1, the program comprising a routine to accept data representing a digitized EEG signal and provide an output representing a numerical value for condensed output information descriptive and/or predictive of cerebral white-matter neural injury in the subject.

24. A method for predicting cerebral white-matter neural injury in a subject comprising acquiring, over a period of time, an EEG signal from the subject, analyzing a frequency distribution of the signal within the range of about 1 Hz to about 50 Hz, and producing condensed output information indicating presence and severity of a cerebral white-matter neural injury.

25. The method of claim 24 further comprising analyzing an intensity of the EEG signal within the range of about 2 Hz to about 20 Hz.

26. The method of claim 25 further comprising repeatedly determining a numerical value for the upper spectral edge of the frequency distribution of the intensity within the EEG signal for each of a series of time intervals, storing a corresponding series of spectral edge values, and presenting the series of spectral edge values in a graphical form.

27. The method of claim 24 further comprising comparing the analyzed data with stored reference spectral edge values and neurological outcome information from subjects of a similar age range and forecasting an outcome useful for managing the subject.

28. The method of claim 27 further comprising the step of rating a subject as follows:

if the spectral edge value is below a first, lowest value, there is a likelihood of severe cerebral white-matter neural injury (WMI);

if the spectral edge value is between the first and a second value, there is a likelihood of moderate WMI;

if the spectral edge value is between the second and a third value, there is a likelihood of mild WMI; and

if the spectral edge value is above the third value, there is little likelihood of WMI.

29. The method of claim 28 where the first value is 6 Hz;

the second value is 8 Hz, and the third value is 10 Hz.

30. The method of claim 27 further comprising the step of rating a subject as follows:

if the spectral edge value, there is a high likelihood of cerebral white-matter neural injury;

if the spectral edge value is between the first and a second value, the subject requires further monitoring; and

if the spectral edge is above the second value, there is a high likelihood that no white-matter injury exists.

31. The method of claim 30 where the first value is 8 Hz and the second value is 10 Hz.

32. The method of claim 24 where the subject is a pre-term infant.

33. The method of claim 24 where the EEG signal is acquired from electrodes placed on the subject'"'"'s head over the parasagittal region/fronto-parietal-occipital cortex.

34. An apparatus for processing an electroencephalogram (EEG) signal from a subject to assist in management of possible cerebral white-matter neural injury in the subject, comprising:

(a) apparatus to acquire at least one channel of an EEG signal from the subject over a period of time, and

(b) computing apparatus programmed to analyze the frequency distribution of the intensity of the EEG signal so acquired, from within a range of from 1 Hz to 50 Hz, and to produce a numerical value for the upper spectral edge of the frequency distribution of the intensity of the EEG signals.

35. The apparatus of claim 34 further comprising apparatus to determine the spectral edge for each of a series of time intervals, to store a corresponding series of spectral edge values, and to include the series of spectral edge values on a display, thereby to enable forecasting of an outcome of cerebral white-matter neural injury in the subject.

36. The apparatus of claim 35 further comprising apparatus to determine, within the range of 1 Hz to 35 Hz, the upper spectral edge below which about 95% of the EEG intensity occurs.

37. The apparatus of claim 36 further comprising apparatus to provide a comparison of the determined upper spectral edge with stored EEG upper spectral edge and associated neurological outcome information from previous cases, thereby to enable forecasting of an outcome of cerebral white-matter neural injury in the subject.

38. The apparatus of claim 37 further comprising apparatus to acquire a plurality of channels of EEG signals, so that spectral edge recordings are made from a plurality of positions on the head of the subject, thereby to improve confidence in forecasting an outcome of cerebral white-matter neural injury in the subject.

39. A program embodied on a computer readable medium in combination with the apparatus of claim 38, the program comprising a routine to display comparable portions of both a current record and a previously collected record thereby permitting a comparison of the current record with stored information derived from the EEG and neurological outcome information from previous cases, thereby enabling prediction of an outcome of cerebral white-matter neural injury in the subject.

40. The apparatus of claim 36 further comprising apparatus programmed to eliminate an artifact in output information by action of at least one event-responsive filtering procedure to delete from the output information data likely to be in error.

41. The apparatus of claim 40 where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of a condition within the EEG signal selected from abnormally low signal amplitude, abnormally high signal amplitude, a signal indicating a seizure in the subject, a signal indicative of sleep in the subject, and a signal indicative of a presence of electrical interference.

42. A program embodied on a computer readable medium in combination with the apparatus of claim 40, the program comprising a routine to display an annotated image of a head to indicate a suitable position for the placement of EEG electrodes.

43. A program embodied on a computer readable medium in combination with the apparatus of claim 41, the program comprising at least one routine to accept data representing a digitized EEG signal and filter the data to exclude data likely to be in error, thereby facilitating prediction of the outcome of cerebral white-matter neural injury in the subject.

44. The program embodied on a computer readable medium of claim 43 further comprising at least one routine to receive information and to act on the data in response to the information to minimize the effect of an artifact.

45. The program embodied on a computer readable medium of claim 43 further comprising at least one routine to analyze the incoming data to establish the likely presence of an artifact, and, if the likely presence of an artifact is established, to output and advisory message to an operator to cause the removal of the artifact.

46. The apparatus of claim 40 further comprising a movement sensor and where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of movement.

47. The apparatus of claim 40 further comprising an electrode impedance detector and where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of an incorrect electrode impedance.

48. The apparatus of claim 40 further comprising an electrode identification detector and where the event-responsive filtering procedure is a signal interruption procedure that rejects the determined upper spectral edge during detection of an inappropriate electrode.

49. The apparatus of claim 40 further comprising an input apparatus through which an operator can input information describing the developmental status of the subject and apparatus programmed to make alterations to filter parameters in accordance with known neurophysiological parameters of subjects of that developmental status.

50. The apparatus of claim 40 further comprising apparatus for issuing an appropriate advisory message to an operator for the operator to cause removal of the artifact.

51. The apparatus of claim 35 further comprising apparatus to determine, between 2 Hz and 20 Hz, the upper spectral edge below which about 90% of the EEG intensity occurs.

52. A method for using the apparatus of claim 35 including screening at least one subject from time to time by using the apparatus to define the characteristics of the EEG of the subject to determine whether the subject being screened is at risk of cerebral white-matter neural injury to allow treatment and or forecasts of likely outcomes to be made accordingly.

53. A method for using the apparatus of claim 35 including continuously monitoring at least one subject by using the apparatus to define and follow the characteristics of the EEG of the subject in order to determine the presence and extent of cerebral white-matter neural injury to allow effectiveness of any treatment of the subject to be monitored and/or to allow a forecast of likely outcome to be made.

54. A program embodied on a computer readable medium in combination with the apparatus of claim 34, the program comprising a routine to accept data representing a digitized EEG signal and provide an output representing a numerical value for the upper spectral edge of the frequency distribution of the intensity within the EEG signal.

57. The method of claim 45 further comprising repeatedly determining a numerical value for the upper spectral edge of the frequency distribution of the intensity within the EEG signal for each of a series of time intervals, storing a corresponding series of spectral edge values, and presenting the series of spectral edge values in a graphical form.

58. The method of claim 45 further comprising comparing the analyzed data with stored reference a spectral edge value and neurological outcome information from one or more subjects of a similar age range and forecasting an outcome useful for managing the subject.

59. The method of claim 58 further comprising the step of rating a subject as follows:

(a) if the spectral edge value is below a first, lowest value, there is a likelihood of severe cerebral white-matter neural injury (WMI);

if the spectral edge value is between the first and a second value, there is a likelihood of moderate WMI;

if the spectral edge value is between the second and a third value, there is a likelihood of mild WMI; and

if the spectral edge value is above the third value, there is little likelihood of WMI.

60. The method of claim 59 where the first value is 6 Hz;

the second value is 8 Hz, and the third value is 10 Hz.

61. The method of claim 58 further comprising the step of rating a subject as follows:

if the spectral edge value, there is a high likelihood of cerebral white-matter neural injury;

if the spectral edge value is between the first and a second value, the subject requires further monitoring; and

if the spectral edge is above the second value, there is a high likelihood that no white-matter injury exists.

62. The method of claim 61 where the first value is 8 Hz and the second value is 10 Hz.

55. A method for predicting cerebral white-matter neural injury in a subject comprising:

(a) acquiring, over a period of time, an EEG signal from the subject,

(b) analyzing a frequency distribution of the signal within the range of about 1 Hz to about 50 Hz,

(c) determining a spectral edge of said frequency distribution; and

(d) comparing the spectral edge determined in step (c) with a spectral edge determined from one or more subjects without cerebral white-matter neural injury to predict cerebral white-matter neural injury.

56. The method of claim 55 further comprising analyzing an intensity of the EEG signal within the range of about 2 Hz to about 20 Hz.

63. The method of claim 55 where the subject is a pre-term infant.

64. The method of claim 55 where the EEG signal is acquired from electrodes placed on the subject'"'"'s head over the parasagittal region/fronto-parietal-occipital cortex.