Method and apparatus for monitoring the condition of a fetus

• US 20030187364A1
• Filed: 04/01/2002
• Published: 10/02/2003
• Est. Priority Date: 04/01/2002
• Status: Active Grant

• Current Assignee: PeriGen, Inc. (Halma PLC)
• Original Assignee: LMS Medical Systems, Ltd. (Halma PLC)
• Inventors: Macieszczak, Maciej, Warrick, Philip A., Hamilton, Emily F., Glaude, Michael C.
• Granted Patent: US 7,113,819 B2
• Time in Patent Office: Days
• US Class Current: 600/511
• CPC Class Codes: A61B 5/0011 Foetal or obstetric data
• CPC Class Codes: A61B 5/02411 of foetuses
• CPC Class Codes: A61B 5/344 Foetal cardiography
• CPC Class Codes: A61B 5/4362 Assessing foetal parameters
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A method and apparatus for monitoring the condition of a fetus to assess a degree of risk of developing a permanent neurological condition is provided. A signal indicative of a fetal heart is processed to derive a degree of risk of developing a permanent neurological condition. The data indicative of the degree of risk of developing a permanent neurological condition indicates a likelihood that the condition of the fetus belongs to a class in a group of classes, where each class in the group of classes is associated with a pre-defined fetal condition. Data indicative of the degree of risk of developing a permanent neurological condition is then released. In one example, a neural network is used to obtain data indicating the likelihood that the condition of the fetus belongs to a class in the group of classes.

Claims

1. An apparatus suitable for monitoring the condition of a fetus, said apparatus comprising:

a) an input for receiving a signal indicative of a fetal heart rate;

b) a processing unit coupled to said input, said processing unit being operative for processing said signal indicative of the fetal heart rate to derive data indicative of a degree of risk of developing a permanent neurological condition, the data indicative of the degree of risk of developing a permanent neurological condition indicating a likelihood that the condition of the fetus belongs to a class in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

c) an output for releasing the data indicative of the degree of risk of developing a permanent neurological condition.

2. An apparatus as defined in claim 1, wherein said input is a first input, said apparatus comprising a second input for receiving a signal indicative of uterine activity, said processing unit being coupled to said first and second inputs and being operative for processing said signal indicative of the fetal heart rate and said signal indicative of uterine activity to derive the data indicative of a degree of risk of developing a permanent neurological condition.

3. An apparatus as defined in claim 1, wherein for each class in the group of classes, the data indicates a likelihood value that the condition of the fetus belongs to the corresponding class.

4. An apparatus as defined in claim 1, wherein said group of classes includes at least four classes comprising:

a) a first class and a second class associated to respective pre-defined fetal conditions indicative of a presence of new-born encephalopathy; and

b) a third class and a fourth class associated to respective pre-defined fetal conditions indicative of an absence of new-born encephalopathy.

5. An apparatus as defined in claim 4, wherein said first class is associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a measure of base deficit above a pre-determined level.

6. An apparatus as defined in claim 4, wherein said second class is associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a measure of base deficit below a pre-determined level.

7. An apparatus as defined in claim 4, wherein said third class is associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level.

8. An apparatus as defined in claim 4, wherein said fourth class is associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

9. An apparatus as defined in claim 1, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a pre-determined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

d) a fourth class associated to a predetermined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

10. An apparatus as defined in claim 1, wherein said processing unit includes a neural network.

11. An apparatus as defined in claim 10, wherein said neural network includes a plurality of inputs and a set of outputs, the outputs corresponding to respective classes in the group of classes, said neural network being adapted to release at each of said outputs a likelihood value associated to a respective class in the group of classes.

12. An apparatus as defined in claim 11, wherein said processing unit includes a pre-processing unit operative to:

a) process the signal indicative of a fetal heart rate to derive a plurality of feature measures;

b) providing said plurality of feature measures to the plurality of inputs of said neural network.

13. An apparatus as defined in claim 12, wherein said signal indicative of a fetal heart rate include heart rate information collected over a certain period of time.

14. An apparatus as defined in claim 13, wherein the certain period of time is in excess of 1 hour.

15. An apparatus as defined in claim 13, wherein the certain period of time is in excess of 2 hour.

16. An apparatus as defined in claim 13, wherein the certain period of time is in excess of 3 hour.

17. An apparatus as defined in claim 13, wherein the certain period of time is in excess of 4 hour.

20. A method as defined in claim 17, wherein for each class in the group of classes, the data indicates a likelihood value that the condition of the fetus belongs to the corresponding class.

21. A method as defined in claim 17, wherein said group of classes includes at least four classes comprising:

a) a first class and a second class associated to respective pre-defined fetal conditions indicative of the presence of new-born encephalopathy; and

b) a third class and a fourth class associated to respective pre-defined fetal conditions indicative of the absence of new-born encephalopathy.

22. A method as defined in claim 21, wherein said first class is associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a measure of base deficit above a pre-determined level.

23. A method as defined in claim 21, wherein said second class is associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a measure of base deficit below a predetermined level.

24. A method as defined in claim 21, wherein said third class is associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level.

25. A method as defined in claim 21, wherein said fourth class is associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

26. A method as defined in claim 17, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a predetermined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a pre-determined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

d) a fourth class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

27. A method as defined in claim 17, wherein said method comprises processing said signal indicative of the fetal heart rate using a neural network to derive data indicative of a degree of risk of developing a permanent neurological condition.

28. A method as defined in claim 27, wherein said neural network includes a plurality of inputs and a set of outputs, the outputs corresponding to respective classes in the group of classes, said neural network being adapted to release at each of said outputs a likelihood value associated to a respective class in the group of classes.

29. A method as defined in claim 28, wherein said method comprises:

a) processing the signal indicative of a fetal heart rate to derive a plurality of feature measures;

b) providing said plurality of feature measures to the plurality of inputs of said neural network.

30. A method as defined in claim 29, wherein said signal indicative of a fetal heart rate include heart rate information collected over a certain period of time.

31. A method as defined in claim 30, wherein the certain period of time is in excess of 1 hour.

32. A method as defined in claim 30, wherein the certain period of time is in excess of 2 hour.

33. A method as defined in claim 30, wherein the certain period of time is in excess of 3 hour.

34. A method as defined in claim 30, wherein the certain period of time is in excess of 4 hour.

18. A method suitable for monitoring the condition of a fetus, said method comprising:

a) receiving a signal indicative of a fetal heart rate;

b) processing said signal indicative of the fetal heart rate to derive data indicative of a degree of risk of developing a permanent neurological condition, the data indicative of the degree of risk of developing a permanent neurological condition indicating a likelihood that the condition of the fetus belongs to a class in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

c) releasing the data indicative of the degree of risk of developing a permanent neurological condition.

19. A method as defined in claim 18, said method further comprising:

a) receiving a signal indicative of uterine activity;

b) processing said signal indicative of the fetal heart rate and said signal indicative of uterine activity to derive the data indicative of a degree of risk of developing a permanent neurological condition.

35. A computer readable storage medium including a program element suitable for execution by a computing apparatus for monitoring the condition of a fetus, said computing apparatus comprising:

a) a memory unit;

b) a processor operatively connected to said memory unit, said program element when executing on said processor being operative for;

i. receiving a signal indicative of a fetal heart rate;

ii. processing said signal indicative of the fetal heart rate to derive data indicative of a degree of risk of developing a permanent neurological condition, the data indicative of the degree of risk of developing a permanent neurological condition indicating a likelihood that the condition of the fetus belongs to a class in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

iii. releasing the data indicative of the degree of risk of developing a permanent neurological condition.

36. A computer readable storage medium as defined in claim 35, wherein said program element when executing on said processor being operative for:

a) receiving a signal indicative of uterine activity;

b) processing said signal indicative of the fetal heart rate and said signal indicative of uterine activity to derive the data indicative of a degree of risk of developing a permanent neurological condition.

37. A computer readable storage medium as defined in claim 35, wherein for each class in the group of classes, the data indicates a likelihood value that the condition of the fetus belongs to the corresponding class.

38. A computer readable storage medium as defined in claim 35, wherein said group of classes includes at least four classes comprising:

a) a first class and a second class associated to respective pre-defined fetal conditions indicative of the presence of new-born encephalopathy; and

b) a third class and a fourth class associated to respective pre-defined fetal conditions indicative of the absence of new-born encephalopathy.

39. A computer readable storage medium as defined in claim 38, wherein said first class is associated to a predetermined fetal condition indicative of the presence of new-born encephalopathy and a measure of base deficit above a pre-determined level.

40. A computer readable storage medium as defined in claim 38, wherein said second class is associated to a predetermined fetal condition indicative of the presence of new-born encephalopathy and a measure of base deficit below a pre-determined level.

41. A computer readable storage medium as defined in claim 38, wherein said third class is associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level.

42. A computer readable storage medium as defined in claim 38, wherein said fourth class is associated to a predetermined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

43. A computer readable storage medium as defined in claim 35, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a pre-determined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

d) a fourth class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

44. A computer readable storage medium as defined in claim 35, wherein said program element when executing on said processor being operative for implementing a neural network, the neural network being adapted to derive data indicative of a degree of risk of developing a permanent neurological condition.

45. A computer readable storage medium as defined in claim 44, wherein said neural network includes a plurality of inputs and a set of outputs, the outputs corresponding to respective classes in the group of classes, said neural network being adapted to release at each of said outputs a likelihood value associated to a respective class in the group of classes.

46. A computer readable storage medium as defined in claim 45, wherein said program element when executing on said processor being operative for:

a) processing the signal indicative of a fetal heart rate to derive a plurality of feature measures;

b) providing said plurality of feature measures to the plurality of inputs of said neural network.

47. A computer readable storage medium as defined in claim 46, wherein said signal indicative of a fetal heart rate include heart rate information collected over a certain period of time.

48. A computer readable storage medium as defined in claim 47, wherein the certain period of time is in excess of 1 hour.

49. A computer readable storage medium as defined in claim 47, wherein the certain period of time is in excess of 2 hour.

50. A computer readable storage medium as defined in claim 47, wherein the certain period of time is in excess of 3 hour.

51. A computer readable storage medium as defined in claim 47, wherein the certain period of time is in excess of 4 hour.

52. A trained neural network comprising:

a) an input unit for receiving feature measures derived from a signal indicative of a fetal heart rate of a fetus; and

b) a set of outputs associated to respective classes in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

c) a processing core coupled to said input unit and to said set of outputs, said processing core being operative to release at each output data indicative of a likelihood that the condition of the fetus belongs to a respective class on the basis of the feature measures received at said input unit.

53. A trained neural network comprising as defined in claim 52, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a predetermined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a pre-determined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a predetermined level;

d) a fourth class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

54. A computer readable storage medium including a program element suitable for execution by a computing apparatus for implementing a trained neural network, said trained neural network comprising:

a) an input unit for receiving feature measures derived from a signal indicative of a fetal heart rate of a fetus;

b) a set of outputs associated to respective classes in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

c) a processing core coupled to said input unit and to said set of outputs, said processing core being operative to release at each output data indicative of a likelihood that the condition of the fetus belongs to a respective class on the basis of the feature measures received at said input unit.

55. A computer readable storage medium as defined in claim 54, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a pre-determined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

d) a fourth class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

56. A fetal monitoring system comprising:

a) a sensor for receiving a signal indicative of a fetal heart rate;

b) an apparatus suitable for monitoring the condition of a fetus, said apparatus comprising;

i. an input coupled to said sensor for receiving a signal indicative of a fetal heart rate;

ii. a processing unit coupled to said input, said processing unit implementing a neural network operative for processing said signal indicative of the fetal heart rate to derive data indicative of a degree of risk of developing a permanent neurological condition, the data indicative of the degree of risk of developing a permanent neurological condition indicating a likelihood that the condition of the fetus belongs to a class in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

iii. an output for releasing the data indicative of the degree of risk of developing a permanent neurological condition;

c) an output unit coupled to the output for said apparatus, said output unit being suitable for displaying the data indicative of the degree of risk of developing a permanent neurological condition.

57. An apparatus suitable for monitoring the condition of a fetus, said apparatus comprising:

a) means for receiving a signal indicative of a fetal heart rate;

b) means for processing said signal indicative of the fetal heart rate to derive data indicative of a degree of risk of developing a permanent neurological condition, the data indicative of the degree of risk of developing a permanent neurological condition indicating a likelihood that the condition of the fetus belongs to a class in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

c) means for releasing the data indicative of the degree of risk of developing a permanent neurological condition.

58. An method for training a neural network suitable for monitoring the condition of a fetus, said method comprising:

a) receiving a plurality of records, each record comprising;

i. a first entry indicative of a fetal heart rate signal;

ii. a second entry indicative of a class selected from a group of classes, each class in said group of classes being indicative of a pre-defined fetal condition;

b) conditioning the neural network on the basis of the first and second entries of each record such as to enable the neural network, upon reception of a signal indicative of a fetal heart rate to derive a likelihood that the condition of the fetus belongs to a class in the group of classes.

59. A method as defined in claim 58, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a predetermined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

d) a fourth class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

60. A method as defined in claim 59, wherein said method comprises conditioning the neural network on the basis of said first and second entries of each record such as to enable the neural network, upon reception of a signal indicative of a fetal heart rate to assign likelihood values to each class in the group of classes, at least in part on the basis of the signal indicative of a fetal heart rate.

61. A method as defined in claim 60, wherein said neural network includes a plurality of inputs and a set of outputs, the outputs corresponding to respective classes in the group of classes, said method comprising:

a) processing the signal indicative of a fetal heart rate to derive a plurality of feature measures;

b) providing said plurality of feature measures to the plurality of inputs of said neural network.

62. A neural network conditioned on the basis of the method described in claim 58.

63. An apparatus suitable for monitoring the condition of a fetus, said apparatus comprising:

a) an input for receiving a signal indicative of a fetal heart rate;

b) a feature extraction unit coupled to said input, said feature extraction unit being operative for processing the signal indicative of the fetal heart rate to derive feature measures;

c) a neural network module coupled to said feature extraction unit, said neural network being operative to process the feature measures to derive data indicative of a degree of risk of developing a permanent neurological condition, the data indicative of the degree of risk of developing a permanent neurological condition indicating a likelihood that the condition of the fetus belongs to a class in a group of classes, each class in the group of classes being associated with a pre-defined fetal condition;

d) an output for releasing the data indicative of the degree of risk of developing a permanent neurological condition.

64. An apparatus as defined in claim 63, wherein said feature extraction unit is operative for:

a) processing said fetal heart rate to identifying a plurality of feature events;

b) generating a plurality of feature measures for the plurality of feature events identified in a);

c) summarising said plurality of feature measures to generate a compact representation of feature measures;

d) providing the compact representation of feature measures to said neural network module.

65. An apparatus as defined in claim 64, wherein summarising said plurality of feature measures to generate a compact representation of feature measures includes:

a) processing the plurality of feature measures to generate at least one interpolated feature measure signal;

b) decimating the interpolated feature measure signal such as to generate the compact representation of feature measures.

66. An apparatus as defined in claim 65, wherein decimating the interpolated feature measure signal is effected on the basis of regular sampling intervals.

67. An apparatus as defined in claim 65, wherein decimating the interpolated feature measure signal is effected on the basis of irregular sampling intervals.

68. An apparatus as defined in claim 64, wherein the plurality of feature events includes at least some events selected from the set consisting of baseline, acceleration, deceleration and contraction events.

69. An apparatus as defined in claim 68, wherein subsets of the plurality of feature measures are associated to respective feature events, at least one subset being associated to a baseline feature event, said at least one subset including feature measures selected from the set consisting of mean value, frequency, length, variability and sinusoidal pattern.

70. An apparatus as defined in claim 68, wherein subsets of the plurality of feature measures are associated to respective feature events, at least one subset being associated to an acceleration feature event, said at least one subset including feature measures selected from the set consisting of mean value, frequency, length, variability and area.

71. An apparatus as defined in claim 68, wherein subsets of the plurality of feature measures are associated to respective feature events, at least one subset being associated to a contraction feature event, said at least one subset including a feature measure selected from the set consisting of frequency and deceleration recovery time.

72. An apparatus as defined in claim 64, wherein said group of classes includes at least four classes comprising:

a) a first class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

b) a second class associated to a pre-determined fetal condition indicative of the presence of new-born encephalopathy and a base deficit measurement below a pre-determined level;

c) a third class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement above a pre-determined level;

d) a fourth class associated to a pre-determined fetal condition indicative of the absence of new-born encephalopathy and a base deficit measurement below a pre-determined level.

73. An apparatus as defined in claim 63, wherein said neural network module includes a plurality of inputs and a set of outputs, the outputs corresponding to respective classes in the group of classes, said neural network module being adapted to release at each of said outputs a likelihood value associated to a respective class in the group of classes.

74. An apparatus as defined in claim 73, wherein the compact representation of feature measures is provided to the plurality of inputs of said neural network module.