Posture state classification for a medical device

• US 9,545,518 B2
• Filed: 04/30/2009
• Issued: 01/17/2017
• Est. Priority Date: 07/11/2008
• Status: Active Grant

• Current Assignee: Medtronic Incorporated (Medtronic Plc)
• Original Assignee: Medtronic Incorporated (Medtronic Plc)
• Inventors: Panken, Eric J., Skelton, Dennis M.
• Primary Examiner(s): Moran, Edward
• Assistant Examiner(s): MAHMOOD, NADIA AHMAD
• Application Number:
  US12/432,993
• Publication Number:
  US 20100010583A1
• Time in Patent Office: 2,819 Days
• US Class Current: 1/1
• CPC Class Codes: A61B 2562/0219 Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
• CPC Class Codes: A61B 5/1116 Determining posture transitions
• CPC Class Codes: A61B 5/1118 Determining activity level
• CPC Class Codes: A61B 5/686 Permanently implanted devices, e.g. pacemakers, other stimulators, biochips A61B5/6861 takes precedence
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61N 1/36535 controlled by body position or posture
• CPC Class Codes: A61N 1/36542 controlled by body motion, e.g. acceleration
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

Techniques for posture classification of a patient in a coordinate system of a sensor. According to one aspect, a defined vector is obtained from a sensor disposed in a substantially fixed manner relative to the patient. The defined vector is described in a coordinate system of the sensor and without regard to an orientation in which the sensor is disposed in relation to the patient. A detected vector is obtained from the sensor that is described using the coordinate system of the sensor. The detected vector and the defined vector to are used to classify the posture state of the patient without regard to the orientation in which the sensor is disposed in relation to the patient. A response may be initiated by a medical device, which may include adjusting therapy delivery.

Claims

1. A method of classifying a posture state of a patient, comprising:

obtaining a defined vector from at least one sensor disposed in a substantially fixed manner relative to the patient, the defined vector being described in a coordinate system of the at least one sensor and without regard to an orientation in which the sensor is disposed in relation to the patient;

obtaining a detected vector from the at least one sensor, the detected vector being described in the coordinate system of the at least one sensor and being indicative of the posture state of the patient;

comparing the detected vector and the defined vector;

classifying the posture state of the patient based on the comparison and without regard to the orientation in which the sensor is disposed in relation to the patient; and

initiating via a medical device an action related to providing care for the patient, the action being based on the posture state classification.

2. The method of claim 1, wherein obtaining the defined vector includes:

measuring outputs of the at least one sensor while the patient assumes a defined posture state; and

associating the measured outputs with the defined vector.

3. The method of claim 1, further comprising associating a region in space with the defined vector, and wherein comparing the detected vector and the defined vector further comprises determining whether the detected vector lies within the region in space.

4. The method of claim 3, further comprising:

obtaining multiple defined vectors, each being associated with a respective region in space; and

wherein comparing the detected vector and the defined vector comprises determining whether the detected vector lies within any of the regions in space associated with the multiple defined vectors.

5. The method of claim 1, further comprising:

associating a tolerance with the defined vector; and

determining a similarity between the defined vector and the detected vector; and

wherein comparing the detected vector and the defined vector comprises comparing the similarity to the tolerance.

6. The method of claim 5, wherein the similarity is a distance, and further comprising:

mapping the distance to a value that is directly proportional to a degree of similarity between the defined vector and the detected vector; and

wherein comparing the detected vector and the defined vector comprises comparing the value to the tolerance.

7. The method of claim 1, further comprising:

obtaining multiple defined vectors;

associating each of the multiple defined vectors with a respective tolerance;

interrelating multiple ones of the tolerances; and

wherein comparing the detected vector and the defined vector comprises comparing the detected vector to a region in space described by the interrelated multiple ones of the tolerances.

8. The method of claim 7, wherein interrelating multiple ones of the tolerances comprises using one or more logical functions.

9. The method of claim 7, wherein the region in space is defined by a toroid and at least one cone.

10. The method of claim 1, further comprising:

associating multiple tolerances with the defined vector, each indicating a relationship to the defined vector;

evaluating a condition;

selecting one of the multiple tolerances based on the evaluation; and

wherein comparing the detected vector and the defined vector comprises determining whether the detected vector satisfies a relationship to the defined vector indicated by the selected one of the multiple tolerances.

11. The method of claim 10, further comprising:

obtaining at least one other defined vector; and

wherein evaluating the condition comprises comparing the detected vector to the at least one other defined vector.

12. The method of claim 11, further comprising:

determining whether the detected vector is closer to a defined vector for a Face Up posture or a defined vector for a Face Down posture; and

selecting a size of a region in space disposed about a defined vector for a Lying Down posture based on the determination; and

wherein comparing the detected vector and the defined vector comprises comparing the detected vector to the region in space to determine whether the patient is in the Lying Down posture.

13. The method of claim 1, further comprising:

deriving one or more constants based on at least one of a squared length of the defined vector and a size of a cone surrounding the defined vector;

associating the one or more constants with the defined vector; and

using the one or more constants to compare the detected vector and the defined vector.

14. The method of claim 13, further comprising normalizing the defined vector to have a selected length.

15. The method of claim 1, further comprising:

obtaining multiple defined vectors;

comparing the detected vector to ones of the multiple defined vectors; and

if, based on the comparison, the patient is not classified as being in a posture state associated with any of the ones of the multiple defined vectors, generating refinement information to identify to which of the ones of the multiple defined vectors the detected vector is closest.

16. The method of claim 1, further comprising:

obtaining multiple defined vectors;

processing ones of the multiple defined vectors to obtain a virtual vector; and

comparing the detected vector and the virtual vector to classify the posture state of the patient.

17. The method of claim 16, wherein comparing the detected vector and the defined vector determines whether the patient is in an Upright posture, and further comprising comparing the detected vector to the virtual vector to determine whether the patient is in a Lying Down posture.

18. The method of claim 17, further comprising:

obtaining multiple additional defined vectors, each being associated with a posture that may be assumed when the patient is prone; and

if the patient is classified as being in the Lying Down posture, comparing the detected vector to one or more of the additional defined vectors to classify the patient as being in one of the postures that may be assumed when the patient is prone.

19. The method of claim 16, wherein comparing the detected vector to the virtual vector comprises determining whether the detected vector falls within any one of multiple regions of space disposed about the virtual vector.

20. The method of claim 16, wherein ones of the multiple defined vectors are each associated with a respective posture the patient may assume when the patient is prone;

obtaining a cross-product between multiple adjacent pairs of the ones of the multiple defined vectors; and

averaging the cross-products to obtain the virtual vector.

21. The method of claim 1, further comprising:

obtaining multiple defined vectors;

classifying the posture state of the patient based on a comparison between the detected vector and each of the multiple defined vectors; and

if the patient is not classified as being in a posture state associated with any of the multiple defined vectors, generating posture refinement information to identify to which of the multiple defined vectors the detected vector is closest.

22. The method of claim 1, further comprising:

obtaining multiple defined vectors;

deriving an inner product between the detected vector and multiple ones of the defined vectors; and

based on the derived inner products, classifying the patient as being in a posture state associated with the one of the multiple ones of the defined vectors to which the detected vector is closest.

23. The method of claim 1, wherein the defined vector and the detected vector are each associated with a direction of motion of the patient, and wherein classifying the posture state of the patient comprises classifying motion of the patient.

24. The method of claim 1, wherein the action comprises adjusting a manner in which therapy is being delivered to the patient.

25. An implantable medical device to determine a posture state of a patient, comprising:

a sensor configured to be substantially fixed to a patient without regard to sensor orientation and to provide signals indicative of a detected vector;

a storage device configured to store at least one posture state definition referencing at least one defined vector expressed in a coordinate system of the sensor; and

a processor configured to obtain the detected vector in the coordinate system of the sensor, to compare the detected vector to the at least one defined vector and to classify a posture state of the patient based on the comparison and without regard to sensor orientation.

26. The device of claim 25, wherein the at least one defined vector describes motion of the patient, and the processor is configured to classify the posture state of the patient based on the motion.

27. The device of claim 25, further comprising a therapy module configured to provide therapy to the patient based on the classification of the posture state.

28. The device of claim 27, wherein the therapy module delivers electrical stimulation therapy to the patient based on the classification of the posture state.

29. The device of claim 25, wherein the at least one posture state definition references a tolerance describing a relationship to the defined vector, and the processor is configured to classify the posture state based on whether the detected vector has the relationship to the at least one defined vector.

30. The device of claim 25, wherein the processor is configured to classify the posture state based on a similarity between the detected vector and each of the at least one defined vector, the similarity being derived without derivation of an angle.

31. The device of claim 25, wherein the processor is configured to classify the posture state based on a determination as to which of the at least one defined vector the detected vector is closest.

32. The device of claim 25, wherein the at least one posture state definition takes into account at least one of posture and activity state of the patient.

33. The device of claim 25, wherein the at least one posture state definition references multiple defined vectors, and the processor is configured to process ones of the multiple defined vectors to obtain a virtual vector, and to classify the posture state of the patient based on a comparison between the detected vector and the virtual vector.

34. The device of claim 33, wherein the processor is configured to obtain multiple adjacent pairs of the ones of the multiple defined vectors, to obtain a cross-product for each of the multiple adjacent pairs, and to average the cross-products to obtain the virtual vector.

35. The device of claim 34, wherein the processor is configured to compare the detected vector to a defined vector for an Upright posture to determine whether the posture state involves an Upright posture, and if not, to compare the detected vector to the virtual vector to determine whether the posture state involves a Lying Down posture.

36. The device of claim 25, wherein the at least one posture state definition references multiple defined vectors, each being associated with a respective tolerance, and wherein a function describes a relationship between the tolerances, and the processor is configured to determine whether the detected vector lies within a region of space described by the relationship between the tolerances.

37. The device of claim 36, wherein at least one of the tolerances describes a toroid surrounding a defined vector for an Upright posture, a different one of the tolerances describes a cone surrounding a defined vector for a posture assumed when a patient is in a prone position, and the relationship is a logical OR function.

38. A method, comprising:

receiving a defined vector from at least one sensor carried in any orientation within a patient, the defined vector being expressed in a coordinate system of the at least one sensor and without regard to the coordinate system of the patient;

defining a posture state in reference to the defined vector;

obtaining a detected vector from the at least one sensor communicatively coupled to an implantable medical device, the detected vector being expressed in the coordinate system of the at least one sensor and without regard to the coordinate system of the patient;

classifying the posture state of the patient based on a comparison between the detected vector and the defined vector; and

delivering therapy to the patient via the implantable medical device based on the posture state classification.

39. The method of claim 38, wherein defining the posture state comprises defining an activity state describing motion of the patient, and wherein the determining step determines whether the patient is engaged in the motion.

40. The method of claim 38, wherein the defined vector is one of a velocity vector and an acceleration vector.

41. The method of claim 38, further comprising:

receiving multiple defined vectors from the at least one sensor;

selecting a respective tolerance for each of the multiple defined vectors;

defining the posture state in reference to ones of the multiple defined vectors and the respective tolerances; and

wherein classifying the posture state of the patient comprises determining whether the detected vector falls within a region in space defined by the posture state definition.

42. The method of claim 38, further comprising:

associating the posture state definition with multiple tolerances and a condition; and

wherein classifying the posture state of the patient comprises evaluating the condition to select one of the multiple tolerances and performing the comparison based on the selected one of the multiple tolerances.

43. The method of claim 39, wherein classifying the posture state of the patient comprises deriving a similarity between the detected vector and the defined vector for use in the comparison between the detected vector and the defined vector.

44. The method of claim 43, wherein the similarity is a distance, and further comprising mapping the distance to a value between 0 and a selected maximum similarity value.

45. A non-transitory storage medium storing executable instructions to cause a processor to perform the method, comprising:

obtaining a detected vector from signals of a sensor, wherein the sensor is carried in any fixed orientation relative to a patient, and wherein the detected vector is indicated in a coordinate system of the sensor;

obtaining one or more posture state definitions that reference one or more defined vectors, each defined vector being indicated in the coordinate system of the sensor;

comparing the detected vector directly to the one or more defined vectors; and

based on the comparison, classifying a posture state of the patient.

46. A system, comprising:

a sensor disposed in any substantially fixed position relative to a patient and configured to provide signals indicative of a detected vector; and

one or more processors configured to obtain a defined vector expressed in a coordinate system of a sensor wherein the defined vector is associated with a posture state definition, to obtain the detected vector in the coordinate system of the sensor, to compare the detected vector to the defined vector, and to classify a posture state of the patient based on the comparison and without regard to the orientation in which the sensor is disposed in relation to the patient.

47. The system of claim 46, further comprising a device external to the patient that comprises one of the one or more processors.

48. The system of claim 47, wherein the device external to the patient is a programmer.

49. The system of claim 48, wherein the programmer includes a user interface to provide information about the posture state of the patient.

50. The system of claim 47, further comprising an implantable medical device comprising one of the one or more processors, and wherein the implantable medical device is configured to communicate with the device external to the patient.

51. The system of claim 46, further comprising an implantable medical device comprising one of the one or more processors.

52. The system of claim 51, wherein the implantable medical device is configured to deliver therapy to the patient based on the classification of the posture state of the patient.

53. The system of claim 46, wherein the at least one defined vector describes motion of the patient, and the processor is configured to classify the posture state of the patient based, at least in part, on the motion.

54. The system of claim 46, wherein the defined vector is associated with a tolerance that describes a relationship to the defined vector, and the processor is configured to classify the posture state based on whether the detected vector has the relationship to the at least one defined vector.

55. The system of claim 46, wherein the processor is configured to classify the posture state based on a similarity between the detected vector and each of the at least one defined vector, the similarity being derived without derivation of an angle.