IMPLANTABLE MEDICAL DEVICE WITH ADAPTIVE SIGNAL PROCESSING AND ARTIFACT CANCELLATION

• US 20100179444A1
• Filed: 01/15/2010
• Published: 07/15/2010
• Est. Priority Date: 01/15/2009
• Status: Active Grant

• Current Assignee: Medtronic Incorporated (Medtronic Plc)
• Original Assignee: Medtronic Incorporated (Medtronic Plc)
• Inventors: Anderson, David A., Deno, D. Curtis, O'Brien, Richard J., Bloem, David R.
• Granted Patent: US 8,706,202 B2
• Time in Patent Office: Days
• US Class Current: 600/509
• CPC Class Codes: A61B 5/0205 Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
• CPC Class Codes: A61B 5/1455 using optical sensors, e.g. spectral photometrical oximeters
• CPC Class Codes: A61B 5/7207 of noise induced by motion artifacts
• CPC Class Codes: A61N 1/3718 Monitoring of or protection against external electromagnetic fields or currents
• CPC Class Codes: A61N 1/3925 Monitoring; Protecting
• CPC Class Codes: G06F 18/2135 based on approximation criteria, e.g. principal component analysis
• CPC Class Codes: G06F 2218/02 Preprocessing
• CPC Class Codes: G06V 40/15 Biometric patterns based on physiological signals, e.g. heartbeat, blood flow

Abstract

A medical device having a sensor sensing an n-dimensional signal during a first known variable condition and during a second known variable condition different from the first known variable condition, a processor performing principal component analysis (PCA) on the sensed n-dimensional signal to generate a first template corresponding to a principal component of variation associated with the first known variable condition and a second template corresponding to a principal component of variation associated with the second known variable condition, a storage device storing the first template and the second template, and a controller detecting a patient condition in response to the stored templates.

Claims

1. A medical device, comprising:

a sensor sensing an n-dimensional signal during a first known variable condition and during a second known variable condition different from the first known variable condition;

a processor performing principal component analysis (PCA) on the sensed n-dimensional signal to generate a first template corresponding to a principal component of variation associated with the first known variable condition and a second template corresponding to a principal component of variation associated with the second known variable condition;

a storage device storing the first template and the second template; and

a controller detecting a patient condition in response to the stored templates.

2. The device of claim 1, wherein the processor generates one or more principal components orthogonal to the principal component corresponding to the first known variable condition, and determines a reduced multi-dimensional signal sensed during the first known variable condition having less than n-dimensions.

3. The device of claim 2, wherein the processor generates one or more principal components orthogonal to the principal component corresponding to the second known variable condition, and determines a reduced multi-dimensional signal sensed during the second known variable condition having less than n-dimensions.

4. The device of claim 1, wherein the first known variable condition corresponds to pathological condition and the second known variable condition corresponds to a non-pathological condition.

5. The device of claim 1, wherein both the first known variable condition and the second known variable condition correspond to a non-pathological condition.

6. The device of claim 1, wherein the n-dimensional signal corresponds to a signal having one of different signal dimensions, different signal frequencies, different sensor positions within a body, different operating frequencies of the sensor or sensors, or other signal aspects separable according to time, frequency, space or sensor type.

7. The device of claim 1, wherein the processor computes a convergence metric to determine when signal data has converged within an acceptable range of variation for generating the first principal component and the second principal component and terminates the PCA on the sensed n-dimensional signal and stores the templates in response to the computed convergence metric.

8. The device of claim 1, wherein the first known variable condition and the second known variable condition correspond to one of multiple states for a given variable and different combinations of variables.

9. The device of claim 1, wherein, the processor selects, in response to the performed PCA, signal dimensions of the n-dimensional signal to be utilized by the controller during the detecting.

10. The device of claim 9, wherein the processor disables a signal dimension of the n-dimensional signal in response to the signal dimension not contributing to variation of the n-dimensional signal of the principal component.

11. The device of claim 1, further comprising an electrode sensing a cardiac signal that is received by the controller, wherein, in response to detecting a predetermined arrhythmia in response to the received sensed cardiac signal, the controller monitors subsequent n-dimensional signals sensed by the sensor to generate an n-dimensional monitoring signal, determines a compensated signal having fewer than n-dimensions using PCA in response to the monitoring signal and one or both of the templates, and confirms detecting of the predetermined arrhythmia in response to the compensated signal.

12. A method of generating a template in a medical device, comprising:

sensing an n-dimensional signal during a first known variable condition;

sensing an n-dimensional signal during a second known variable condition different from the first known variable condition;

performing principal component analysis (PCA) on the sensed n-dimensional signal sensed for the first known variable condition to generate a first template of a principal component of variation associated with the first known variable condition;

performing principal component analysis (PCA) on the sensed n-dimensional signal sensed for the second known variable condition to generate a second template of a principal component of variation associated with the second known variable condition; and

storing the first template and the second template and detecting a patient condition in response to the stored templates.

13. The method of claim 12, wherein performing principal component analysis (PCA) on the sensed n-dimensional signal sensed for the first known variable condition further comprises:

generating one or more principal components orthogonal to the principal component corresponding to the first known variable condition; and

reducing the n-dimensional signal sensed during the first known variable condition to an m-dimensional signal in response to the generated one or more orthogonal principal components, wherein m is less than n.

14. The method of claim 13, wherein performing principal component analysis (PCA) on the sensed n-dimensional signal sensed for the second known variable condition further comprises:

generating one or more principal components orthogonal to the principal component corresponding to the second known variable condition; and

reducing the n-dimensional signal sensed during the second known variable signal to an m-dimensional signal in response to the one or more orthogonal principal components, wherein m is less than n.

15. The method of claim 12, wherein the first known variable condition corresponds to pathological condition and the second known variable condition corresponds to a non-pathological condition.

16. The method of claim 12, wherein both the first known variable condition and the second known variable condition correspond to a non-pathological condition.

17. The method of claim 12, wherein the n-dimensional signal corresponds to a signal having one of different signal dimensions, different signal frequencies, different sensor positions within a body, different operating frequencies of the sensor or sensors, or other signal aspects separable according to time, frequency, space or sensor type.

18. The method of claim 12, further comprising computing a convergence metric to determine when signal data has converged within an acceptable range of variation for generating the first principal component and the second principal component.

19. The method of claim 12, wherein the first known variable condition and the second known variable condition correspond to one of multiple states for a given variable and different combinations of variables.

20. The method of claim 12, further comprising selecting, in response to the performed PCA, signal dimensions of the n-dimensional signal to be utilized during the detecting.

21. The method of claim 20, further comprising disabling a signal dimension of the n-dimensional signal in response to the signal dimension not contributing to variation of the n-dimensional signal of the principal component.

22. The method of claim 12, further comprising:

sensing a cardiac signal via a sensing electrode;

detecting a predetermined arrhythmia in response to the sensed cardiac signal;

monitoring subsequently sensed n-dimensional signals, in response to detecting the arrhythmia, to generate an n-dimensional monitoring signal;

determining a compensated signal having fewer than n-dimensions using PCA in response to the monitoring signal and one or both of the templates; and

confirming the detection of the predetermined arrhythmia in response to the compensated signal.