METHOD AND APPARATUS FOR DETECTION OF NERVOUS SYSTEM DISORDERS

• US 20070249953A1
• Filed: 12/12/2006
• Published: 10/25/2007
• Est. Priority Date: 04/21/2006
• Status: Abandoned Application

• Current Assignee: Medtronic Incorporated (Medtronic Plc)
• Original Assignee: Medtronic Incorporated (Medtronic Plc)
• Inventors: Frei, Mark, Carlson, David, Drew, Touby, Panken, Eric, Osorio, Ivan
• Application Number:
  US11/609,388
• Publication Number:
  US 20070249953A1
• Time in Patent Office: Days
• US Class Current: 600/544
• CPC Class Codes: A61N 1/36082 Cognitive or psychiatric applications, e.g. dementia or Alzheimer's disease

Abstract

Systems and methods for detecting and/or treating nervous system disorders, such as seizures, are disclosed. Certain embodiments of the invention relate generally to implantable medical devices (IMDs) adapted to detect and treat nervous system disorders in patients with an IMD. Certain embodiments of the invention include detection of seizures based upon comparisons of long-term and short-term representations of physiological signals. Other embodiments include prediction of seizure activity based upon analysis of physiological signal levels. A further embodiment of the invention includes detection of seizure activity following the delivery of therapy.

Claims

1. A method of detecting a neurological event, comprising:

acquiring EEG signal data comprising a stream of data values;

determining a short-term representation of the EEG signal data based on data values acquired over a first sample interval, the short-term representation being determined by applying the data values to a first filter, the first filter adapted to produce an intermediate output value for each block of N data values, and applying a rolling window of M intermediate output values to a second filter, the second filter adapted to produce the short-term representation;

determining a long-term representation of the EEG signal data based on data values acquired over a second sample interval, the second sample interval being longer than the first sample interval;

calculating a ratio of the short-term representation to the long-term representation;

comparing the ratio to an onset threshold; and

detecting a neurological event when the ratio exceeds the onset threshold.

2. The method of claim 1 wherein the neurological event is an epileptic seizure.

3. The method of claim 1 wherein the long-term representation is determined by applying a stream of input values derived from the data values to a third filter, the third filter adapted to produce an intermediate output value for each block of X input values, and applying a rolling window of Y intermediate output values to a fourth filter, the fourth filter adapted to produce the long-term representation.

4. The method of claim 3 wherein the stream of input values comprise the short-term representation values.

5. The method of claim 4 wherein the short-term representation values are downsampled prior to being applied to the third filter.

6. The method of claim 5 wherein the short-term representation values are downsampled by a downsampling factor of 2.

7. The method of claim 3 further comprising setting the value of the long-term representation to a specified minimum value if the value determined using the third and fourth filters is less than the specified minimum value.

8. The method of claim 7 further comprising applying a low-pass filter to the value of the long-term representation.

9. The method of claim 1 wherein determining the ratio of the short-term representation to the long-term representation comprises estimating the ratio as a function of 2 raised to an exponent, the exponent comprising a difference between a most significant set bit (MSSB) position of the long-term representation and a MSSB position of the short-term representation, wherein the MSSB position is a numbered bit position of a first non-zero bit, starting from a most significant bit (MSB).

10. The method of claim 1 wherein the intermediate output value produced by the first filter is a statistical measure of the data values in a given block.

11. The method of claim 10 wherein the statistical measure is a median value.

12. The method of claim 10 wherein the statistical measure is a mean value.

13. The method of claim 1 wherein the short-term representation produced by the second filter is a statistical measure of the data values in the rolling window.

14. The method of claim 13 wherein the statistical measure is a median value.

15. The method of claim 13 wherein the statistical measure is a mean value.

16. The method of claim 1 wherein the first sample interval for determining the short-term representation is a period of less than about 5 minutes.

17. The method of claim 16 wherein the first sample interval for determining the short-term representation is a period of about 2 minutes.

18. The method of claim 1 wherein the second sample interval for determining the long-term representation is a period of greater than about 15 minutes.

19. The method of claim 18 wherein the second sample interval for determining the long-term representation is a period of greater than about 20 minutes.

20. The method of claim 18 wherein the second sample interval for determining the long-term representation is a period of about 30 minutes.

21. The method of claim 3 wherein the third and fourth filters used to determine the long-term representation produce statistical measures of the data values applied to each filter.

22. The method of claim 1 wherein a neurological event is detected when the ratio exceeds the onset threshold for an onset duration.

23. The method of claim 22 wherein a neurological event is detected when a specified number of consecutive values of the ratio exceeds the onset threshold for an onset duration.

24. The method of claim 22 wherein a neurological event is detected when a predetermined percentage of values of the ratio exceeds the onset threshold for a detection duration.

25. The method of claim 1 further comprising identifying an end of a neurological event when the ratio decreases below a termination threshold.

26. The method of claim 25 wherein the end of a neurological event is identified when the ratio decreases below a termination threshold for a termination duration.

27. The method of claim 26 wherein the end of a neurological event is identified when a specified number of consecutive values of the ratio decreases below the termination threshold for a termination duration.

28. The method of claim 26 wherein the end of a neurological event is identified when a predetermined percentage of values of the ratio decreases below the termination threshold for a termination duration.

29. The method of claim 1 wherein the onset threshold is greater than 5.

30. The method of claim 29 wherein the onset threshold is between about 10 and 50.

31. The method of claim 30 wherein the onset threshold is about 22.

32. The method of claim 1 wherein the number of data values N in each block is a specified number in a range from 3 to 25 data values.

33. The method of claim 32 wherein the number of data values N in each block is a specified number in a range from 11 to 21 data values.

34. The method of claim 33 wherein the N data values in each block includes 15 successive data values.

35. The method of claim 1 wherein the number of intermediate output values M in a rolling window is a specified number in a range from 3 to 101 intermediate output values.

36. The method of claim 35 wherein the number of intermediate output values M in a rolling window is a specified number in a range from 11 to 51 intermediate output values.

37. The method of claim 36 wherein the intermediate output values M in a rolling window include 29 to 35 successive intermediate output values.

38. The method of claim 35 wherein the M intermediate output values are representative of a period of less than about 5 minutes.

39. The method of claim 38 wherein the M intermediate output values are representative of a period of about 2 minutes.

40. The method of claim 3 wherein the number of data values X in each block is a specified number in a range from 3 to 25 data values.

41. The method of claim 40 wherein the number of data values X in each block is a specified number in a range from 11 to 21 data values.

42. The method of claim 41 wherein the X data values in each block includes 15 successive data values.

43. The method of claim 3 wherein the number of intermediate output values Y in a rolling window is a specified number in a range from 3 to 101 intermediate output values.

44. The method of claim 43 wherein the number of intermediate output values Y in a rolling window is a specified number in a range from 11 to 51 intermediate output values.

45. The method of claim 44 wherein the intermediate output values Y in a rolling window include 29 to 35 successive intermediate output values.

46. The method of claim 43 wherein the Y intermediate output values are representative of a period of greater than about 15 minutes.

47. The method of claim 46 wherein the Y intermediate output values are representative of a period of greater than about 20 minutes.

48. The method of claim 47 wherein the Y intermediate output values are representative of a period of about 30 minutes.

49. A computer-readable medium programmed with instructions for performing a method of detecting a neurological event, the medium comprising instructions for causing a programmable processor to:

acquire EEG signal data comprising a stream of data values;

determine a short-term representation of the EEG signal data based on data values acquired over a first sample interval, the short-term representation being determined by applying the data values to a first filter, the first filter adapted to produce an intermediate output value for each block of N data values, and applying a rolling window of M intermediate output values to a second filter, the second filter adapted to produce the short-term representation;

determine a long-term representation of the EEG signal data based on data values acquired over a second sample interval, the second sample interval being longer than the first sample interval;

calculate a ratio of the short-term representation to the long-term representation;

compare the ratio to an onset threshold; and

detect a neurological event when the ratio exceeds the onset threshold.

50. The medium of claim 49 further comprising instructions to determine the long-term representation by applying the short-term representation values to a third filter, the third filter adapted to produce an intermediate output value for each block of X input values, and applying a rolling window of Y intermediate output values to a fourth filter, the fourth filter adapted to produce the long-term representation.

51. The medium of claim 49 further comprising instructions to determine the ratio of the short-term representation to the long-term representation by estimating the ratio as a function of 2 raised to an exponent, the exponent comprising a difference between a most significant set bit (MSSB) position of the long-term representation and a MSSB position of the short-term representation, wherein the MSSB position is a numbered bit position of a first non-zero bit, starting from a most significant bit (MSB).

52. An implantable medical device system for detecting a neurological event, the system comprising:

an implantable medical device (IMD); and

at least one electrode adapted to sense EEG signals from a brain of a patient and communicate the EEG signals to the device,

wherein the device is adapted to

acquire EEG signal data comprising a stream of data values;

determine a short-term representation of the EEG signal data based on data values acquired over a first sample interval, the short-term representation being determined by applying the data values to a first filter, the first filter adapted to produce an intermediate output value for each block of N data values, and applying a rolling window of M intermediate output values to a second filter, the second filter adapted to produce the short-term representation;

determine a long-term representation of the EEG signal data based on data values acquired over a second sample interval, the second sample interval being longer than the first sample interval;

calculate a ratio of the short-term representation to the long-term representation;

compare the ratio to an onset threshold; and

detect a neurological event when the ratio exceeds the onset threshold.

53. The system of claim 52 comprising at least two electrodes adapted to sense at least two channels of EEG signals, wherein the IMD is adapted to calculate the ratio of the short-term representation to the long-term representation for the at least two channels, and compare the at least two ratios to an onset threshold.

54. The system of claim 53 wherein the IMD is adapted to detect a neurological event when the ratio of at least one channel exceeds the onset threshold.

55. The system of claim 53 wherein the IMD is adapted to detect a neurological event when a maximum ratio value exceeds the onset threshold, the maximum ratio value equal to a maximum of the ratio values.

56. The system of claim 52 wherein the IMD is further adapted to deliver therapy after a neurological event is detected.

57. The system of claim 56 wherein the IMD is adapted to deliver drug therapy.

58. The system of claim 56 wherein the IMD is adapted to deliver electrical stimulation therapy.

59. The system of claim 58 wherein the IMD is adapted to deliver electrical stimulation therapy via at least one electrode also used for sensing EEG signals.

60. The system of claim 58 wherein the IMD is adapted to begin charging electronic circuitry when a neurological event is detected.